FDA 510(k) Database: How to Search, Find Predicates, and Use Clearance Data

What Is the FDA 510(k) Database?

The FDA 510(k) database is the publicly searchable repository of every premarket notification (510(k)) submission the FDA has received and acted upon since the Medical Device Amendments of 1976. It is maintained by the Center for Devices and Radiological Health (CDRH) and contains records for over 120,000 cleared devices — making it the single largest source of regulatory intelligence for medical devices sold in the United States.

Every time a manufacturer submits a 510(k) to demonstrate that their device is substantially equivalent to a legally marketed predicate device, the FDA logs the submission in this database. Once a decision is made — whether the device is found substantially equivalent (SE), not substantially equivalent (NSE), or the submission is withdrawn — the record becomes publicly accessible.

This database is not optional reading for regulatory professionals. It is the primary tool for:

• Finding predicate devices for new 510(k) submissions
• Researching competitor products and their regulatory clearance history
• Understanding what the FDA has accepted in terms of testing approaches, intended use statements, and technological comparisons
• Tracking regulatory trends across device categories over time
• Validating product classification and confirming which product codes apply to your device type

If you are preparing a 510(k) submission, predicate research in this database is where your regulatory strategy begins. If you are conducting competitive intelligence, market analysis, or due diligence for an acquisition, this database is an indispensable resource.

Key distinction: The 510(k) database contains clearance records — administrative and decision data about each submission. It is not the same as the Product Classification Database (which defines device classes and product codes) or MAUDE (which tracks adverse events). These databases complement each other, and effective regulatory research requires using all of them together.

Legal and Regulatory Foundation

The 510(k) process is rooted in Section 510(k) of the Federal Food, Drug, and Cosmetic (FD&C) Act, with implementing regulations codified in 21 CFR Part 807, Subpart E. Understanding this legal framework is essential for interpreting what the database records actually represent.

What Is Substantial Equivalence?

Substantial equivalence (SE) is the legal standard that underpins every 510(k) clearance in the database. A device is substantially equivalent to a predicate if it meets one of two criteria:

1. Same intended use AND same technological characteristics as the predicate device, OR
2. Same intended use AND different technological characteristics, but the differences do not raise new questions of safety and effectiveness — and the submitter demonstrates that the device is at least as safe and effective as the predicate

Substantial equivalence does not mean identical. The FD&C Act explicitly permits differences in design, materials, energy source, and other characteristics, provided the submitter demonstrates those differences do not compromise safety or effectiveness.

The FDA evaluates SE with respect to: intended use, design, energy used or delivered, materials, performance, safety, effectiveness, labeling, biocompatibility, standards, and other applicable characteristics.

The SE Determination Process

When the FDA reviews a 510(k), it follows a structured decision process:

1. Acceptance review (Refuse to Accept screening) — Within 15 calendar days, the FDA checks whether the submission is administratively complete. If critical elements are missing, the FDA issues a Refuse to Accept (RTA) decision, and the submission is returned without substantive review. The current RTA rate is approximately 10% to 15%.
2. Intended use comparison — Does the new device have the same intended use as the predicate? If no, the device is not substantially equivalent.
3. Technological characteristics comparison — Does the new device have the same technological characteristics as the predicate? If yes, and performance data demonstrates equivalence, the device is SE.
4. Different technological characteristics evaluation — If the technology differs, do the differences raise new questions of safety or effectiveness? If no, and performance data demonstrates equivalence, the device is SE. If the differences raise new questions that cannot be resolved, the device is found not substantially equivalent (NSE).

This decision logic is documented in the FDA's guidance: "The 510(k) Program: Evaluating Substantial Equivalence in Premarket Notifications [510(k)]" (2014).

Why this matters for database research: When you review 510(k) summaries in the database, you are reading the applicant's SE argument — how they framed their device's intended use, how they compared technological characteristics to the predicate, and what performance data they presented. Understanding the SE framework helps you evaluate the strength of a predicate candidate and anticipate what the FDA will expect from your own submission.

Preamendments vs. Postamendments Devices

The 510(k) database primarily contains postamendments devices — devices marketed after the Medical Device Amendments of May 28, 1976. However, predicate devices can also be preamendments devices — devices that were legally marketed before that date.

Preamendments devices (also called "grandfathered" devices) were legally on the market before May 28, 1976, and were not:

• Significantly changed or modified after that date
• Subject to a PMA requirement

These devices do not have 510(k) numbers because they predated the 510(k) requirement. If you use a preamendments device as a predicate, you may need to provide documentation establishing its preamendment status. However, because medical technology has advanced significantly since 1976, the FDA recommends using a recently cleared postamendments device as your predicate whenever possible.

Postamendments devices are the vast majority of entries in the 510(k) database — devices cleared through the 510(k) process after 1976. Almost all modern 510(k) submissions cite postamendments predicates.

Who Must Submit a 510(k)?

The database records submissions from four categories of submitters:

1. US medical device manufacturers — The most common submitters. Any company manufacturing a Class II (or certain Class I/III) device for the US market must submit a 510(k) before marketing.
2. US representatives of foreign manufacturers — Non-US manufacturers must have a US-based representative (initial importer or US Agent) submit the 510(k) on their behalf. The database records the US representative as the applicant.
3. Specification developers — Companies that design and develop a device but outsource manufacturing to a contract manufacturer (CMO) are responsible for the 510(k), not the CMO.
4. Repackers and relabelers — In rare cases, companies that significantly repack or relabel a device (e.g., adding new intended uses or warnings) may need to submit a 510(k).

The Three Types of 510(k) Submissions

The database records three types of 510(k) submissions, each representing a different level of review:

Type	When Used	Key Characteristics	Typical Review Time
Traditional	Most first-time submissions and devices with significant differences from predicates	Full submission with complete device description, predicate comparison, and performance data. Reviewed entirely by the FDA.	~90 to 130 FDA days
Special	Modifications to a manufacturer's own previously cleared device	Abbreviated format. Only the modification and its impact on safety/effectiveness are reviewed. Must be submitted by the same manufacturer that holds the original clearance.	~30 to 60 FDA days
Abbreviated	Devices that conform to recognized consensus standards or FDA guidance	Relies on declarations of conformity to recognized standards (e.g., IEC 60601-1, ISO 10993) rather than providing full test reports. Summary reports reference the standards.	~60 to 90 FDA days

Understanding these types is important when reading database records. A device cleared via Special 510(k) typically represents an incremental modification, while a Traditional 510(k) may represent a new market entrant or a device with more significant technological differences from the predicate.

Changes That Trigger a New 510(k)

Not every modification to a cleared device requires a new 510(k). The database grows because manufacturers must submit a new 510(k) when they make changes that could significantly affect safety or effectiveness. The FDA's guidance "Deciding When to Submit a 510(k) for a Change to an Existing Device" (2017) outlines when a new submission is required:

• Changes in intended use — Any new or expanded indication for use
• Design changes — Modifications to the device's fundamental technology, materials, energy source, or operating principle that could affect safety or effectiveness
• Manufacturing process changes — Changes in sterilization methods, manufacturing location (in some cases), or processes that affect device performance
• Software changes — Major software revisions that alter the device's core functionality, risk profile, or intended use (minor bug fixes and usability improvements typically do not trigger a new 510(k))
• Labeling changes — Adding new warnings, contraindications, or use instructions that change the device's risk profile

Changes that do not significantly affect safety or effectiveness can be documented internally in a Letter to File rather than a new 510(k) submission. The manufacturer's quality system should include a procedure for evaluating whether a change triggers a new 510(k).

How to Access the FDA 510(k) Database

The FDA provides multiple ways to access 510(k) data. Each serves a different use case.

Primary Web Interface

The main search portal is the 510(k) Premarket Notification database on the FDA's AccessData platform:

URL: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm

This is the interface most regulatory professionals use for day-to-day searches. It provides a form-based search with multiple fields, and results link directly to 510(k) summary documents when available.

FDA 510(k) Clearances Landing Page

The FDA also maintains a broader landing page for 510(k) clearances that provides context, links to downloadable files, and annual clearance lists:

URL: https://www.fda.gov/medical-devices/device-approvals-and-clearances/510k-clearances

This page includes:

• Links to the searchable database
• Downloadable 510(k) files (bulk data)
• Lists of cleared devices organized by year
• Guidance on the 510(k) program

openFDA API

For programmatic access, the openFDA platform provides a RESTful API for 510(k) data:

Base endpoint: https://api.fda.gov/device/510k.json

Documentation: https://open.fda.gov/apis/device/510k/

The API returns data in JSON format and supports complex queries, filtering, counting, and pagination. It is free to use, requires no authentication for basic queries (though an API key increases rate limits), and is the best option for bulk analysis, trend tracking, and integration into regulatory intelligence tools.

Bulk Data Downloads

The FDA provides downloadable ZIP files containing all releasable 510(k) records for specific time periods:

URL: https://www.fda.gov/medical-devices/510k-clearances/downloadable-510k-files

These files are updated monthly, typically on the 5th of each month, and contain tab-delimited data suitable for import into Excel, databases, or analysis tools. Use these when you need to analyze the entire dataset or build your own local database.

Additionally, the complete dataset is available through:

• Data.gov: https://catalog.data.gov/dataset/premarket-notifications-510ks
• openFDA Downloads: https://open.fda.gov/apis/device/510k/download/ — JSON-formatted bulk download files

Access Summary

Access Method	Best For	Format	Update Frequency
AccessData web interface	Individual searches, predicate research	Web form / HTML	Weekly
FDA clearances page	Annual lists, general orientation	Web page / PDF	As published
openFDA API	Programmatic queries, trend analysis, automation	JSON	Quarterly (with rolling updates)
Bulk downloads (FDA)	Full dataset analysis, local database	Tab-delimited text (ZIP)	Monthly
Bulk downloads (openFDA)	Full dataset in JSON, data science workflows	JSON (ZIP)	Quarterly

Step-by-Step Search Tutorial

The AccessData web interface is the primary tool for most regulatory professionals. Here is how to use it effectively.

Understanding the Search Form

When you navigate to the 510(k) Premarket Notification database, you will see a search form with the following fields:

• 510(k) Number — The unique identifier assigned to each submission (e.g., K212345)
• Applicant — The name of the company or individual that submitted the 510(k)
• Device Name — The trade name or common name of the device
• Product Code — The three-letter FDA product code (e.g., QKQ for pulse oximeters, DXN for orthopedic screws)
• Date of Decision — The date range for the FDA's clearance decision
• Decision — The outcome (SESE = substantially equivalent, SEKD = SE with conditions, DENG = denied, etc.)
• Clearance Type — Traditional, Special, or Abbreviated
• Third Party Review — Whether the submission was reviewed by an accredited third party (A3P)
• Expedited Review — Whether the submission received expedited review
• Advisory Committee — The FDA review panel responsible for the device category

Additionally, the FDA has added sub-category filters including:

• Predetermined Change Control Plan (PCCP) Authorized — For devices cleared with a PCCP under the 510(k) pathway
• Clinical Trials — Whether the clearance involved clinical data
• Combination Product — Whether the device is a combination product

Search Method 1: By Product Code (Recommended Starting Point)

Searching by product code is the most reliable and productive method for predicate research. A product code groups all devices of the same type, so you get a comprehensive list of every 510(k) clearance in your device category.

How to do it:

1. Go to the FDA Product Classification Database to find your product code if you do not already know it. Search by device name or description.
2. Return to the 510(k) database.
3. Enter the three-letter product code in the Product Code field (e.g., "DQA" for powered surgical instruments).
4. Leave all other fields blank.
5. Click Search.

The results will show every 510(k) submission for that product code, sorted by decision date. This gives you the entire clearance history for your device type — every competitor, every predicate, and every technology variation the FDA has reviewed.

Why product code first? Device names vary wildly across manufacturers. One company calls it a "Patient Monitor," another calls it a "Vital Signs Monitoring System," and a third uses a brand name with no descriptive content. Product codes cut through this inconsistency because every device in the same classification gets the same code.

Search Method 2: By Applicant Name

Use this when you want to see all 510(k) clearances held by a specific company — for competitive intelligence, due diligence, or to find a specific manufacturer's predicate chain.

How to do it:

1. Enter the company name in the Applicant field. Use the formal legal name as it appears on FDA registrations (e.g., "Medtronic" not "Medtronic plc" or "Medtronic, Inc." — though you may need to try variations).
2. Leave all other fields blank.
3. Click Search.

Tips:

• Company names change through acquisitions and mergers. A device originally cleared by "Covidien" may now be marketed by "Medtronic" following the 2015 acquisition — but the database records reflect the applicant name at the time of submission.
• Try partial names. Entering "Boston" will return results for "Boston Scientific Corporation" and any other applicants with "Boston" in the name.
• If a company has many clearances, combine the applicant name with a product code to narrow results.

Search Method 3: By Device Name

Use this when you know the trade name or common name of a device but not its product code or K-number.

How to do it:

1. Enter a descriptive keyword in the Device Name field.
2. Leave all other fields blank.
3. Click Search.

Critical warning: The device name search matches exact text strings. This means:

• Searching for "infusion pump" will return devices with that exact phrase, but will miss devices listed as "volumetric pump" or "IV pump" or "drug delivery pump."
• Use a single broad keyword rather than a multi-word phrase. "Pump" returns more results than "infusion pump."
• Never rely solely on device name searches for predicate research. Always confirm your findings with a product code search.

Search Method 4: By 510(k) Number

Use this when you already know the specific K-number you are looking for — for example, when checking a competitor's predicate or verifying a reference in published literature.

How to do it:

1. Enter the K-number in the 510(k) Number field. The format is "K" followed by six digits (e.g., K212345). You can also enter just the numeric portion.
2. Click Search.

This returns the single record for that submission.

Search Method 5: By Decision Date Range

Use this to find recently cleared devices in a category, track the pace of clearances over time, or identify seasonal patterns.

How to do it:

1. Enter a date range in the Date of Decision fields (From and To dates).
2. Optionally combine with a product code or advisory committee filter.
3. Click Search.

This is particularly useful when you want to see what the FDA has cleared in the past 12 months for a specific device type — the most recent clearances often reflect the most current FDA expectations for testing and documentation.

Search Method 6: By Advisory Committee

Every device is assigned to one of the FDA's advisory committees (review panels) based on its medical specialty area. Searching by advisory committee returns all 510(k) clearances reviewed by that panel, which is useful for understanding the regulatory landscape within an entire medical specialty.

Common advisory committees include:

Advisory Committee Code	Specialty Area
AN	Anesthesiology
CV	Cardiovascular
CH	Clinical Chemistry
DE	Dental
EN	Ear, Nose, and Throat
GU	Gastroenterology/Urology
HE	Hematology
HO	General Hospital
IM	Immunology
MI	Microbiology
NE	Neurology
OB	Obstetrics/Gynecology
OP	Ophthalmic
OR	Orthopedic
PA	Pathology
PM	Physical Medicine
RA	Radiology
SU	General/Plastic Surgery
TX	Toxicology

Combining Search Fields

You can fill in multiple fields simultaneously to narrow your results. However, there is an important caveat: the database performs an AND operation across all filled fields, and it uses exact matching. This means:

• If you enter both a product code and an applicant name, you get only results that match both.
• If you enter a device name and a product code, but the device name text does not appear in the database record, you get zero results — even if relevant clearances exist under that product code.

Best practice: Start with a single field (preferably product code), review the results, and then add additional fields only to narrow a large result set. Do not start with multiple fields filled in.

Understanding Search Results

Results List

Search results are displayed in a table with the following columns:

Column	Description
510(k) Number	The unique K-number (clickable link to the full record)
Applicant	Company name of the submitter
Device Name	Trade name or common name as entered by the applicant
Product Code	Three-letter product classification code
Date Received	Date the FDA received the submission
Decision Date	Date the FDA issued its decision
Decision	The outcome — SESE (substantially equivalent), SEKD (SE with conditions), DENG (denied/NSE), etc.
Type	Traditional, Special, or Abbreviated

The Full 510(k) Record

Clicking on a 510(k) number opens the detailed record, which includes:

• K-Number and applicant information — Full contact details including address
• Correspondent — The regulatory contact person for the submission
• Device classification information — Product code, regulation number, advisory committee, device class, and review panel
• 510(k) submission type — Traditional, Special, or Abbreviated
• Clearance type — Whether the device was cleared through the standard pathway, third-party review, or other mechanism
• Decision — Substantially equivalent (SE), not substantially equivalent (NSE), or withdrawn
• Predicate devices — The K-numbers of the predicate device(s) cited in the submission (when available in newer records)
• Statement or Summary — A link to the 510(k) Summary document (if the applicant provided a summary rather than a statement)

Decision Codes Explained

Code	Meaning	What It Tells You
SESE	Substantially Equivalent	Standard SE determination — the device was cleared
SEKD	Substantially Equivalent with Conditions	Cleared, but with specific conditions (e.g., post-market commitments)
SESD	Substantially Equivalent — Special Device	SE determination for a special 510(k) device
SESP	Substantially Equivalent — Split Predicate	Cleared using a split predicate approach (intended use from one predicate, technology from another)
DENG	Denied (Not Substantially Equivalent)	The FDA determined the device is NSE — the 510(k) was rejected
UNKN	Unknown	Older records without a coded decision

How to Read a 510(k) Summary

The 510(k) summary is one of the most valuable documents in the database. It is a publicly available document prepared by the submitter that describes the basis for the FDA's substantial equivalence determination. Understanding what the summary contains — and what it does not — is critical for predicate research and competitive intelligence.

What Is Required in a 510(k) Summary

Under 21 CFR 807.92, a 510(k) summary must contain:

1. Submitter identification — Name, address, telephone number, contact person, and date the summary was prepared
2. Device identification — The name of the device, including trade/proprietary name, common/usual name, and classification name
3. Predicate device identification — The legally marketed device(s) to which equivalence is claimed, including the predicate's 510(k) number
4. Device description — How the device functions, the scientific concepts underlying it, and significant physical and performance characteristics (design, materials, physical properties)
5. Intended use statement — The specific medical application or purpose of the device
6. Comparison to the predicate — If the device has the same technological characteristics as the predicate, a summary comparison. If different, an explanation of how the differences do not raise new questions of safety or effectiveness
7. Performance data summary — A brief discussion of nonclinical tests (bench testing, biocompatibility, electrical safety, software verification) and clinical tests (if any), including results related to safety and effectiveness

What Is NOT in the Summary

The 510(k) summary is a condensed document. It does not typically include:

• Full test reports — You get summaries of testing, not the raw data or complete test protocols
• Detailed design specifications — The device description is general, not a blueprint
• Manufacturing process details — How the device is made is proprietary and not disclosed
• Risk analysis — The full risk management file (ISO 14971) is not published
• Software architecture details — For software-containing devices, the summary describes functionality, not source code or detailed architecture documentation
• Proprietary comparison data — Detailed side-by-side specifications beyond what is needed to establish SE

Summary vs. Statement

Submitters have a choice under 21 CFR 807.93: instead of providing a summary, they can file a 510(k) statement — a certification that the submitter will make safety and effectiveness information available to any person within 30 days of a written request. A statement contains almost no technical information.

When you see "Statement" instead of "Summary" in the database, it means the submitter opted not to publish a summary. To obtain the underlying information, you would need to either (a) contact the company directly citing the 510(k) statement obligation, or (b) submit a Freedom of Information Act (FOIA) request to the FDA.

Practical tip: For predicate research, always prioritize devices that have published summaries over those with only statements. A summary lets you confirm that the predicate's intended use and technological characteristics align with yours before you commit to using it. A statement tells you nothing.

Reading the Summary Strategically

When reviewing a 510(k) summary for predicate research, focus on these elements:

1. Intended use statement — Does this exactly match (or encompass) your device's intended use? Any deviation means you may need to justify the difference or find a different predicate.
2. Device description — What materials, design features, and functional characteristics does the predicate use? How do they compare to your device?
3. Predicate chain — What predicate did this device cite? You can trace the chain backward to understand the full regulatory lineage.
4. Testing summary — What types of testing did the FDA review? This gives you a preview of what the FDA will expect for your submission.
5. Standards referenced — What consensus standards (ISO, IEC, ASTM) were cited? These likely form the baseline testing expectations for your device type.

Using the Database for Predicate Device Research

Predicate selection is the most important strategic decision in the 510(k) process. The database is your primary tool for this research. Here is a systematic workflow.

Step 1: Identify Your Product Code

Before you touch the 510(k) database, determine your device's product code using the FDA Product Classification Database. The product code is the foundation of your search because it groups all devices of the same type under a single classification.

Go to: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfPCD/classification.cfm

Search by device name, medical specialty, or regulation number. Note the three-letter product code, the device class, the regulation number, and any special controls listed.

Step 2: Search by Product Code in the 510(k) Database

Enter your product code in the 510(k) database. This returns every 510(k) clearance ever issued for that device type. Sort by decision date (newest first) to start with the most recently cleared devices.

Step 3: Build a Predicate Candidate List

From the results, identify 5 to 10 potential predicate devices based on:

• Recency — Recent clearances reflect current FDA expectations. A predicate cleared in 2025 is more informative than one cleared in 2005.
• Same intended use — Your primary predicate must have the same intended use or your intended use must be a subset of the predicate's.
• Similar technology — Devices with the same technological characteristics simplify your SE argument.
• Available summary — Devices with published summaries let you evaluate the predicate before committing.
• Clean regulatory history — Check whether the device has any recalls or adverse event patterns in MAUDE.

Step 4: Download and Review 510(k) Summaries

For each predicate candidate, download the 510(k) summary and systematically compare:

Comparison Element	Your Device	Predicate Candidate
Intended use / Indications for use	[Your statement]	[Predicate's statement]
Device description	[Your description]	[Predicate's description]
Technological characteristics	[Your technology]	[Predicate's technology]
Materials	[Your materials]	[Predicate's materials]
Performance specifications	[Your specs]	[Predicate's specs]
Testing approach	[Planned]	[What FDA reviewed]
Standards cited	[Your standards list]	[Predicate's standards list]

Step 5: Trace the Predicate Chain

Every 510(k) clearance cites one or more predicates, which in turn cite earlier predicates. Tracing this chain backward accomplishes several things:

• Confirms there is an unbroken chain of legal marketing back to a pre-amendment device or De Novo classification
• Reveals how the FDA's expectations have evolved over time for your device type
• Identifies the specific testing approaches that have been accepted at different points in time
• Exposes any weak links — predicates that were recalled, found NSE on subsequent review, or have significant adverse event histories

To trace the chain, look for predicate K-numbers in each summary document, then search for those K-numbers in the database and repeat the process.

Step 6: Validate with the FDA's Guidance

The FDA has published specific guidance on predicate selection: "Best Practices for Selecting a Predicate Device to Support a Premarket Notification [510(k)] Submission" (2023). This guidance recommends:

• Use a recently cleared device as your predicate
• Avoid predicates with unmitigated safety issues, design-related recalls, or use-related problems
• Document your predicate selection rationale
• Consider a Pre-Submission (Q-Sub) meeting with the FDA if your predicate strategy is complex

Practical tip: If you are unsure about your predicate selection, the Pre-Submission program lets you ask the FDA directly. Submit a Q-Sub request with your proposed predicate(s), intended use statement, and comparison rationale. The FDA will provide feedback before you invest in testing — potentially saving months of work and significant expense.

Advanced Search Strategies

Strategy 1: Reverse-Engineering a Competitor's Regulatory Path

To understand exactly how a competitor brought their device to market:

1. Search by applicant name to find all their clearances
2. Identify the specific device you are interested in
3. Download the 510(k) summary
4. Note the predicate they cited
5. Search for the predicate's summary
6. Repeat — trace the full chain
7. Cross-reference with the Product Classification Database for regulation numbers and special controls
8. Check MAUDE for adverse event history on both the competitor's device and the predicate

This gives you a complete picture of the regulatory precedent your competitor relied on, the testing approaches that were accepted, and any post-market issues that have emerged.

Strategy 2: Identifying All Devices in a Specific Niche

Sometimes you need to understand the full competitive landscape for a narrow device type:

1. Search by product code (primary approach)
2. Supplement with device name keyword searches using different terminology (the same device type may be described with different names by different manufacturers)
3. Filter by date range to focus on the current market (devices cleared in the last 5 to 10 years are most relevant)
4. Export the results and organize by applicant to see market share distribution

Strategy 3: Finding Predicates When Product Codes Do Not Align Perfectly

Your device may bridge two product codes or represent a combination that does not fit neatly into existing classifications. In these cases:

1. Search multiple product codes that could apply to your device
2. Look for "split predicate" clearances (SESP decision code) where the FDA accepted different aspects from different predicates
3. Search by device name keywords that describe your device's function rather than its form factor
4. Review FDA guidance documents for your device type — they often discuss classification and predicate selection considerations
5. Consider a Pre-Submission meeting to confirm the appropriate product code and predicate approach

Strategy 4: Monitoring a Device Category Over Time

To track how FDA expectations are evolving:

1. Search by product code with no date restriction to get the complete history
2. Compare summaries from different time periods — note changes in testing requirements, standards cited, and the level of detail the FDA accepted
3. Pay attention to Additional Information (AI) letters issued during review — these reveal specific questions the FDA asks, which can shift over time
4. Track the ratio of Traditional vs. Abbreviated vs. Special 510(k) submissions to see whether the FDA is encouraging specific submission types for your device category

Strategy 5: Using Sub-Category Filters

The FDA has added sub-category search filters to the 510(k) database to improve search precision:

• PCCP Authorized — Finds devices cleared with a Predetermined Change Control Plan, which allows certain post-clearance modifications without a new 510(k). This is particularly relevant for AI/ML-enabled devices.
• Clinical Trials — Identifies clearances that included clinical data, helping you understand whether the FDA expects clinical evidence for your device type.
• Combination Product — Filters for devices that combine a medical device with a drug or biologic.
• Redacted FOIA 510(k) — Identifies records obtained through FOIA requests where confidential information has been redacted.

Multiple sub-categories can be searched simultaneously.

Strategy 6: Using FOIA for Missing Summaries

When a device of interest has only a 510(k) statement (no published summary), you can submit a Freedom of Information Act (FOIA) request to obtain the underlying information:

1. Go to the FDA's FOIA request portal
2. Identify the specific K-number(s) you need information for
3. Request the 510(k) summary or the non-confidential portions of the submission
4. Allow 20 to 60 business days for processing (complex requests take longer)
5. The FDA will provide redacted versions with trade secrets and confidential commercial information removed

FOIA requests are free for small requests (fewer than 100 pages). Larger requests may incur processing fees. While slower than direct database access, FOIA is sometimes the only way to obtain technical details about older clearances or submissions where the applicant chose not to publish a summary.

Practical tip: Before filing a FOIA request, try contacting the applicant directly. Under 21 CFR 807.93, a submitter who filed a 510(k) statement is required to make safety and effectiveness information available to any person within 30 days of a written request. This is often faster than the FOIA process.

Strategy 7: Building a Predicate Landscape Map

For complex devices that may reference multiple product codes or where predicate chains are deep, create a visual predicate landscape map:

1. Start with your target product code and pull all clearances
2. For the 10 most recent clearances, download each summary and record the cited predicates
3. Search for each cited predicate and record its predicates in turn
4. Continue until you reach pre-amendment devices or De Novo authorizations
5. Plot the relationships as a network diagram — each node is a K-number, each edge is a predicate citation
6. Color-code nodes by applicant to visualize competitive dynamics
7. Mark any nodes with recalls, MAUDE signals, or predicate eliminations

This map reveals the "regulatory genealogy" of your device category. It shows which devices are the foundational predicates that many others depend on, which companies have the deepest predicate chains, and where potential vulnerabilities exist (e.g., a widely-cited predicate that has been recalled or eliminated).

openFDA API for 510(k) Data

The openFDA API transforms the 510(k) database from a manual search tool into a programmable data source. If you need to analyze clearance trends, build internal regulatory intelligence tools, or automate competitive monitoring, the API is the right tool.

Getting Started

Base endpoint:

https://api.fda.gov/device/510k.json

Basic query format:

https://api.fda.gov/device/510k.json?search=FIELD:VALUE&limit=N

No authentication is required for basic queries. However, registering for a free API key at https://open.fda.gov/apis/authentication/ increases your rate limit from 40 requests per minute to 240 requests per minute.

Key Searchable Fields

Field	Description	Example Value
k_number	The 510(k) number	K212345
applicant	Company name	"Medtronic"
device_name	Trade or common device name	"Patient Monitor"
product_code	Three-letter product code	QKQ
date_received	Date FDA received the submission	20230115 (YYYYMMDD)
decision_date	Date of FDA decision	20230601
decision_code	Decision outcome	SESE
decision_description	Human-readable decision	"Substantially Equivalent"
clearance_type	Submission type	Traditional
advisory_committee	Review panel code	CV (Cardiovascular)
advisory_committee_description	Review panel name	"Cardiovascular"
review_advisory_committee	Review committee assignment	—
third_party	Whether third-party reviewed	Y or N
expedited_review_flag	Whether expedited review was used	Y or N
statement_or_summary	Whether a summary or statement was filed	Summary
type_of_submission	Submission category	Traditional
address_1, address_2, city, state, zip_code, country_code	Applicant address components	—
contact	Contact person name	—

Example API Queries

Find all 510(k) clearances for a specific product code:

https://api.fda.gov/device/510k.json?search=product_code:"QKQ"&limit=100

Find all clearances by a specific applicant:

https://api.fda.gov/device/510k.json?search=applicant:"Medtronic"&limit=100

Find clearances within a date range:

https://api.fda.gov/device/510k.json?search=decision_date:[20240101+TO+20241231]&limit=100

Combine multiple search criteria (AND logic):

https://api.fda.gov/device/510k.json?search=product_code:"QKQ"+AND+decision_code:"SESE"&limit=100

Count clearances by product code:

https://api.fda.gov/device/510k.json?count=product_code.exact

Count clearances by year:

https://api.fda.gov/device/510k.json?count=decision_date&limit=50

Count clearances by advisory committee:

https://api.fda.gov/device/510k.json?count=advisory_committee_description.exact

Understanding API Responses

Every API response contains two sections:

{
  "meta": {
    "disclaimer": "...",
    "terms": "...",
    "license": "...",
    "last_updated": "2026-01-15",
    "results": {
      "skip": 0,
      "limit": 100,
      "total": 1547
    }
  },
  "results": [
    {
      "k_number": "K212345",
      "applicant": "ACME MEDICAL INC.",
      "device_name": "CARDIAC MONITOR SYSTEM",
      "product_code": "DRT",
      "decision_date": "2021-06-15",
      "decision_code": "SESE",
      "clearance_type": "Traditional",
      "advisory_committee": "CV",
      "statement_or_summary": "Summary",
      ...
    }
  ]
}

• meta.results.total tells you how many records match your query (even if only limit records are returned in this response)
• results is the array of matching records
• Use the skip parameter to paginate through large result sets (e.g., &skip=100 for the second page of 100 results)

Practical API Use Cases

Use case 1 — Automated competitive monitoring: Set up a scheduled script that queries for new clearances in your product codes weekly. Compare against your list of known competitors to flag new entrants.

Use case 2 — Predicate network mapping: Query all clearances for your product code, extract predicate K-numbers from the results, and recursively query those predicates to build a complete predicate genealogy.

Use case 3 — Market size estimation: Count total clearances by product code and applicant to estimate how many active manufacturers exist in a device category.

Use case 4 — Regulatory timeline benchmarking: Calculate the average time from date_received to decision_date across your product code to benchmark expected review timelines.

Python Example: Querying 510(k) Data

Here is a basic Python example that queries the openFDA API for all 510(k) clearances under a specific product code and writes the results to a CSV file:

import requests
import csv

API_URL = "https://api.fda.gov/device/510k.json"
PRODUCT_CODE = "QKQ"  # Pulse oximeters
LIMIT = 100

results = []
skip = 0

while True:
    params = {
        "search": f'product_code:"{PRODUCT_CODE}"',
        "limit": LIMIT,
        "skip": skip,
    }
    response = requests.get(API_URL, params=params)
    data = response.json()

    if "results" not in data:
        break

    results.extend(data["results"])
    total = data["meta"]["results"]["total"]

    if skip + LIMIT >= total:
        break
    skip += LIMIT

# Write to CSV
with open("510k_results.csv", "w", newline="") as f:
    writer = csv.writer(f)
    writer.writerow(["K-Number", "Applicant", "Device Name",
                      "Decision Date", "Decision", "Type"])
    for r in results:
        writer.writerow([
            r.get("k_number", ""),
            r.get("applicant", ""),
            r.get("device_name", ""),
            r.get("decision_date", ""),
            r.get("decision_code", ""),
            r.get("clearance_type", ""),
        ])

print(f"Exported {len(results)} records to 510k_results.csv")

This script handles pagination automatically and can be adapted for any search criteria. Replace the PRODUCT_CODE variable with your device's product code, or modify the search parameter to query by applicant, date range, or any other field.

API Limitations

• Maximum of 1,000 records per request (use skip to paginate)
• Maximum skip value of 25,000 — for datasets larger than 26,000 records, use the bulk download files instead
• The count endpoint returns up to 1,000 unique values
• Rate limits: 40 requests/minute without an API key, 240/minute with a key
• Some older records have incomplete data (missing fields or inconsistent formatting)
• The API data is not updated in real-time — there is a lag between when a clearance is issued and when it appears in the API

Related FDA Databases

The 510(k) database does not exist in isolation. Effective regulatory research requires cross-referencing multiple FDA databases. Here are the most important ones and how they relate to 510(k) data.

Product Classification Database

URL: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfPCD/classification.cfm

Contains the classification regulation for every FDA-recognized medical device type. Use this first to determine your product code, device class, regulation number, and applicable special controls. This is the prerequisite to any 510(k) database search.

MAUDE (Manufacturer and User Facility Device Experience)

URL: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfmaude/search.cfm

Contains adverse event reports (Medical Device Reports or MDRs) submitted by manufacturers, importers, device user facilities, and voluntary reporters. The FDA receives several hundred thousand MDRs annually. Use MAUDE to check potential predicates for safety signals before committing to them. A predicate with a pattern of adverse events may still be legally valid, but it invites additional scrutiny from FDA reviewers.

PMA (Premarket Approval) Database

URL: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpma/pma.cfm

Contains records of premarket approval applications for Class III devices. While PMA devices cannot serve as 510(k) predicates, the PMA database is useful for understanding the competitive landscape in high-risk device categories and for identifying where the 510(k) boundary ends and PMA territory begins.

De Novo Database

URL: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/denovo.cfm

Contains records of De Novo classification requests. A device that receives De Novo authorization is classified into Class I or II and can subsequently serve as a predicate for 510(k) submissions. If you are working with a novel device type and cannot find a suitable 510(k) predicate, check whether a De Novo has established a classification regulation for your device category.

Establishment Registration and Device Listing

URL: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfRL/rl.cfm

Contains registration information for manufacturers, importers, and specification developers, along with listing information for the devices they market. Use this to verify that a company is properly registered and that a specific device is actively listed — information that can be relevant for predicate validation and competitive intelligence.

GUDID (Global Unique Device Identification Database)

URL: https://accessgudid.nlm.nih.gov/

Contains device identification information based on the Unique Device Identification (UDI) system. Use GUDID to cross-reference commercial product names with 510(k) numbers — useful when you know a competitor's product name but not their regulatory clearance number.

Recalls Database

URL: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfres/res.cfm

Contains records of medical device recalls. Cross-reference potential predicate devices against the recalls database to check for any Class I, II, or III recalls. The FDA's 2023 predicate selection guidance specifically recommends avoiding predicates with design-related recalls.

Database Cross-Reference Workflow

For comprehensive regulatory research, use these databases in sequence:

1. Product Classification Database — Determine your product code and classification
2. 510(k) Database — Find predicate candidates and competitor clearances
3. De Novo Database — Check for newly classified device types if no 510(k) predicate exists
4. MAUDE — Screen predicate candidates for adverse event patterns
5. Recalls Database — Screen predicate candidates for recall history
6. GUDID — Cross-reference commercial product names with regulatory records
7. Establishment Registration — Verify manufacturer registration status

Common Search Mistakes and How to Avoid Them

Mistake 1: Searching Multiple Fields Simultaneously

Problem: Filling in both a product code and a device name often returns zero results because the device name text must exactly match the database record.

Solution: Search one field at a time. Start with product code, then use the results to identify specific records. Add additional fields only when you have confirmed the exact text the database uses.

Mistake 2: Using Overly Specific Device Names

Problem: Entering "Bluetooth-enabled continuous positive airway pressure device" returns nothing because no record uses that exact phrase.

Solution: Use single generic keywords: "CPAP" or "airway" or "pressure." Then scan the results visually for relevant devices.

Mistake 3: Not Accounting for Company Name Changes

Problem: Searching for "Philips" but missing devices cleared under "Respironics" (before the Philips acquisition) or "Volcano Corporation" (before the Philips Volcano merger).

Solution: Research the corporate history of the company you are investigating. Search for all known previous names, subsidiaries, and acquired companies separately.

Mistake 4: Ignoring Older Records

Problem: Filtering only for recent clearances and missing the foundational predicate that the entire device category is built on.

Solution: Always run at least one search with no date filter to see the full clearance history. Older clearances establish the predicate lineage that current devices depend on.

Mistake 5: Assuming the Database Is Complete

Problem: Treating the 510(k) database as a comprehensive list of all legally marketed devices.

Solution: Remember that the database does not include Class I exempt devices (no 510(k) required), Class II exempt devices, pre-amendment devices that were never the subject of a 510(k), devices marketed under PMA or De Novo pathways, or devices that were legally marketed but later withdrawn. Use it as one source among several, not the only source.

Mistake 6: Confusing "Cleared" with "Currently Marketed"

Problem: Assuming that a device listed in the database is still on the market. A 510(k) clearance does not expire, but the device may have been discontinued, recalled, or superseded by a newer model.

Solution: Cross-reference with the FDA Establishment Registration and Device Listing database to check whether the device is still actively listed. Check GUDID for current device identification information.

Mistake 7: Relying on the Database for Detailed Technical Information

Problem: Expecting the 510(k) record or summary to contain sufficient technical detail to fully characterize a competitor's device or reverse-engineer their design.

Solution: The summary is a regulatory document, not a technical specification. Use it for regulatory intelligence — intended use, technological approach, testing strategy, predicate selection — but not as a substitute for actual product evaluation or teardown analysis.

Using 510(k) Data for Competitive Intelligence

The 510(k) database is one of the most underutilized competitive intelligence resources in the medical device industry. Because every company must disclose their device description, intended use, testing approach, and predicate selection, the database provides a window into competitors' regulatory strategies that is not available in any other industry.

What You Can Learn About Competitors

Intelligence Type	How to Extract It
Product portfolio	Search by applicant name to see every cleared device
Market entry timeline	Compare date received to decision date for each clearance
Regulatory strategy	Analyze predicate selections and 510(k) types (Traditional vs. Abbreviated vs. Special)
Technology evolution	Compare summaries across multiple generations of the same product line
Geographic presence	Applicant address and manufacturing site information
Testing approach	Read summaries for cited standards, test methods, and clinical data
Expansion plans	New clearances in adjacent product codes signal market expansion
Acquisition integration	Post-acquisition clearances show how an acquirer is integrating a target's product line

Competitive Intelligence Workflow

1. Identify your competitive set — List all companies you compete with or may compete with.
2. Map their clearance portfolios — Search by applicant name for each competitor. Download results and organize by product code, clearance date, and predicate.
3. Analyze predicate strategies — Determine which predicates they chose and why. This reveals their positioning relative to other devices in the market.
4. Track new clearance activity — Set up periodic (monthly or quarterly) monitoring of new clearances in your product codes to detect new entrants and product launches.
5. Review 510(k) summaries — Read summaries for testing approaches, standards used, and clinical data. This informs your own testing strategy and identifies potential differentiation points.
6. Cross-reference with MAUDE — Check for adverse event patterns that may indicate product quality issues or emerging safety concerns.

Using 510(k) Data for Market Sizing

The database enables rough market sizing for medical device categories:

• Count unique applicants by product code to estimate the number of active manufacturers
• Track clearance volume over time to assess whether a category is growing, stable, or declining
• Identify geographic concentration from applicant address data
• Estimate product lifecycle by measuring the time between initial clearance and successor clearances from the same applicant

Using 510(k) Data for Regulatory Strategy

Beyond competitive intelligence, the 510(k) database directly informs your regulatory strategy.

Informing Your Submission Approach

• Determine the dominant 510(k) type — If most recent clearances in your product code are Traditional, expect a full submission. If many are Abbreviated, the FDA may accept a standards-based approach.
• Identify expected testing — Summaries reveal what biocompatibility testing, performance testing, electrical safety testing, software validation, and clinical data the FDA has reviewed for similar devices.
• Benchmark review timelines — Calculate the average time from submission to decision for your product code to set realistic timeline expectations.
• Anticipate AI letter questions — If multiple summaries in your category reference similar additional information the FDA requested, you can proactively address those topics in your submission.

Predicate Selection Strategy

The database is the single source for predicate selection. Your strategy should consider:

• Recency: The FDA's 2023 guidance recommends using recently cleared predicates. They reflect current FDA expectations and reduce the risk of the predicate being later eliminated.
• Clean history: Cross-reference with MAUDE and Recalls to avoid predicates with safety signals.
• Summary availability: Predicates with published summaries let you verify the intended use and testing approach before you commit.
• Same manufacturer vs. different manufacturer: If you are modifying your own device, a Special 510(k) referencing your own prior clearance is fastest. If referencing a competitor's device, you need a Traditional or Abbreviated approach.

Pre-Submission Meeting Preparation

When preparing for a Pre-Submission (Q-Sub) meeting with the FDA, the 510(k) database helps you:

• Identify 2 to 3 predicate candidates with clear rationale for each
• Show the FDA you understand the clearance history for your device type
• Demonstrate awareness of how testing expectations have evolved
• Ask informed questions about specific testing approaches the FDA has previously accepted or rejected

510(k) Clearance Statistics

Understanding the scale and trends of the 510(k) program provides context for your regulatory planning.

Historical Volume

The 510(k) program has been in operation since 1976. Key statistics:

• Total clearances to date: Over 120,000 devices have been cleared through the 510(k) pathway
• Annual volume: The FDA clears approximately 3,000 to 3,500 devices per year through the 510(k) pathway
• Percentage of devices: Between 95% and 98% of medical devices on the US market entered through the 510(k) pathway
• Submission types: Approximately 70% are Traditional, 20% are Abbreviated, and 10% are Special 510(k) submissions

Recent Trends

Metric	Approximate Value
Annual 510(k) clearances	~3,000 to 3,500
Average review time (calendar days, median)	~108 to 142 days
Average review time (FDA days)	~90 to 100 days
Refuse to Accept (RTA) rate	~10% to 15%
NSE (denial) rate	<3%
AI/ML device clearances (2025)	295
PCCP-authorized 510(k)s	Growing (new sub-category tracked since 2024)

Key Trends Affecting the Database

• Predicate elimination: Since 1976, the FDA has eliminated 1,758 devices as eligible predicates. 84% of these eliminations (1,477 devices) occurred since 2012, reflecting the FDA's push to modernize the predicate base.
• PCCP growth: The introduction of Predetermined Change Control Plans allows devices (especially AI/ML-enabled devices) to be modified within pre-approved parameters without a new 510(k). This is creating a new class of database records.
• Enhanced sub-categories: The FDA's addition of search sub-categories (PCCP, Clinical Trials, Combination Product) reflects the growing complexity of submissions and the need for better data granularity.
• AI/ML medical devices: The rapid growth of AI/ML-enabled devices — 295 clearances in 2025 alone — is driving significant changes in how the database is used and how predicate networks are structured.

Practical Workflows for Common Use Cases

Workflow 1: "I Need to Find a Predicate for a New 510(k)"

1. Determine your product code via the Product Classification Database
2. Search the 510(k) database by product code (no other filters)
3. Sort results by decision date (newest first)
4. Identify 5 to 10 candidates with SE decisions
5. Download summaries for each candidate
6. Compare intended use statements to your device's intended use
7. Evaluate technological characteristics for similarity
8. Check MAUDE and Recalls for each candidate
9. Select 1 primary predicate and 1 to 2 reference predicates
10. Document your selection rationale per FDA guidance

Workflow 2: "I Need to Understand What Testing the FDA Will Expect"

1. Search by product code for the 10 most recent clearances
2. Download all available 510(k) summaries
3. Create a matrix of testing types across all summaries (biocompatibility, electrical safety, EMC, software, performance, clinical)
4. Identify which consensus standards were cited most frequently
5. Note any clinical data submissions — if multiple recent clearances included clinical data, expect the FDA to require it for yours
6. Review any FDA guidance documents specific to your device type
7. Use the findings to build your testing plan

Workflow 3: "A Competitor Just Launched a New Product — What Can I Learn?"

1. Search by applicant name to find their latest clearance
2. Download the 510(k) summary
3. Note the intended use — has it expanded or narrowed compared to their previous device?
4. Identify the predicate they cited — is it their own prior device (suggesting incremental improvement) or a competitor's device (suggesting a market entry)?
5. Review the testing summary — what new testing did they conduct?
6. Check whether they used a Traditional, Abbreviated, or Special 510(k) — this reveals their regulatory approach
7. Cross-reference with their establishment registration to identify manufacturing locations

Workflow 4: "I Am Evaluating a Company for Acquisition — What Is Their Regulatory Portfolio?"

1. Search by applicant name (and all known predecessor names) in the 510(k) database
2. Search the PMA and De Novo databases as well
3. Map all clearances by product code, date, and type
4. Identify the predicate chain for each active product — are predicates owned by the target or by third parties?
5. Check MAUDE for adverse event history on each product
6. Check the Recalls database for any recall history
7. Verify establishment registrations and device listings are current
8. Assess whether clearances are transferable or whether new 510(k)s would be needed post-acquisition

Frequently Asked Questions

Is the FDA 510(k) database free to access?

Yes. The 510(k) database, including the web search interface, bulk downloads, and the openFDA API, is entirely free and open to the public. No account, subscription, or login is required. The API offers higher rate limits with a free API key.

How often is the 510(k) database updated?

The web-based database (AccessData) is updated weekly with newly reviewed 510(k) notifications. Downloadable bulk data files are updated monthly, typically on the 5th of each month. The openFDA API is updated quarterly with rolling updates in between. There is always a lag between when a clearance decision is made and when the record appears in the database.

Can I find 510(k) submissions that were denied (found NSE)?

Yes. The database includes all submissions the FDA has acted upon, including those that received a Not Substantially Equivalent (NSE) determination (decision code DENG). You can filter for these by selecting the appropriate decision code in the search form or using decision_code:"DENG" in the openFDA API. However, denied submissions typically do not have published summaries.

What is the difference between a 510(k) summary and a 510(k) statement?

A 510(k) summary is a publicly available document prepared by the submitter that describes the device, predicate, comparison, and testing approach — it provides substantive technical and regulatory information. A 510(k) statement (21 CFR 807.93) is simply a certification by the submitter that they will make safety and effectiveness information available upon written request within 30 days. A statement contains virtually no technical information. Summaries are far more useful for research purposes.

How do I find the predicate device for a specific 510(k)?

For recent clearances, the predicate K-number(s) may be listed in the detailed record. For all clearances, download the 510(k) summary — it will identify the predicate device by name and K-number. If only a statement is available, you can submit a FOIA request to obtain the predicate information, or contact the applicant directly.

Can I use a device that has been recalled as my predicate?

Yes — a recalled device can still serve as a legally valid predicate, provided it was at one point legally marketed. However, the FDA's 2023 guidance on predicate selection recommends against using predicates with design-related recalls, as this invites additional scrutiny and questions from reviewers. Use a recalled predicate only if there is a compelling reason and no suitable alternative exists.

Does a 510(k) clearance expire?

No. A 510(k) clearance does not expire. Once the FDA finds a device substantially equivalent, that determination remains in effect indefinitely. However, the device must continue to comply with all applicable regulations, and the manufacturer must maintain current establishment registration and device listing. The clearance applies to the specific device as described in the 510(k) — significant modifications require a new 510(k).

How many 510(k)s are cleared each year?

The FDA clears approximately 3,000 to 3,500 devices per year through the 510(k) pathway. The exact number varies year to year based on submission volume, FDA staffing, and review complexity. The median review time has ranged from approximately 108 to 142 calendar days in recent years.

Can I use the openFDA API without an API key?

Yes. The openFDA API works without authentication. However, unauthenticated requests are rate-limited to 40 per minute and 1,000 per day. Registering for a free API key increases these limits to 240 per minute and 120,000 per day — which is necessary for any bulk analysis or automated monitoring.

What does "Third Party Review" mean in the 510(k) database?

Third Party Review (also called Accredited Persons or A3P review) means the 510(k) was reviewed by an FDA-accredited third party organization rather than by the FDA directly. The FDA makes the final clearance decision, but the technical review is conducted by the third party. This option is available for certain lower-risk Class II devices and can result in faster review times. The FDA maintains a list of accredited third-party review organizations at accessdata.fda.gov. To be eligible for third-party review, a device must generally be a low-to-moderate-risk Class II device that is not intended to be permanently implanted, not life-sustaining or life-supporting, and not for a use subject to an FDA recall within the past two years.

What happens to my clearance if my predicate is eliminated?

If the FDA eliminates a predicate device (removes it from eligibility as a predicate), your existing clearance remains valid. A predicate elimination does not retroactively revoke clearances that relied on that predicate. However, if you plan to submit a new 510(k) that would have cited the eliminated predicate, you will need to identify an alternative predicate. The FDA has eliminated over 1,750 devices as eligible predicates since 1976, with approximately 84% of those eliminations occurring since 2012.

What is the difference between 510(k) clearance and PMA approval?

A 510(k) results in clearance — a determination that the device is substantially equivalent to a legally marketed predicate. This is different from PMA approval, which is a more rigorous process requiring independent evidence of safety and effectiveness (typically including clinical trials) for Class III high-risk devices. The 510(k) database contains clearances only; PMA records are in a separate database. A PMA-approved device cannot serve as a predicate for a 510(k) submission unless it has been reclassified (downclassified) from Class III to Class II or Class I.

Can a device cleared under De Novo serve as a predicate?

Yes. Once a device receives De Novo authorization and is classified into Class I or II, it can serve as a predicate device for subsequent 510(k) submissions. De Novo authorizations effectively create new device classifications, opening the 510(k) pathway for follow-on devices. Check the De Novo database to see if a recently authorized device has established a new classification that applies to your product.

Third-Party Commercial Search Tools

While the FDA's free AccessData interface and openFDA API are the authoritative sources for 510(k) data, several commercial platforms offer enhanced search capabilities, data visualization, and workflow integration that go beyond what the raw government databases provide.

What Commercial Tools Offer

Commercial 510(k) search and regulatory intelligence platforms typically provide:

• Enhanced search interfaces — More intuitive search with auto-complete, synonym matching, and fuzzy text search that compensates for the inconsistencies in device naming that make the FDA's exact-match search difficult to use
• Predicate chain visualization — Automated mapping of predicate relationships as network diagrams, showing the complete regulatory genealogy for a device category without manual K-number-by-K-number tracing
• Automated monitoring — Alerts when new clearances, recalls, or adverse events appear in your product codes or for your competitors
• Cross-database integration — Unified search across the 510(k) database, MAUDE, Recalls, Product Classification, GUDID, and standards databases in a single interface
• Collaboration features — Shared workspaces for regulatory teams to annotate and compare predicate candidates, track submission progress, and document predicate selection rationale
• Analytics and trend reports — Pre-built dashboards for clearance volumes, review times, submission types, and testing requirements by product code

When to Use Commercial Tools vs. FDA Sources

Scenario	Recommended Approach
Quick lookup of a single K-number or product code	FDA AccessData (free, authoritative)
Deep predicate research for a submission	Commercial tool for efficiency, then verify against FDA source
Automated competitive monitoring across multiple product codes	Commercial tool or custom openFDA API scripts
Bulk data analysis or academic research	openFDA API or bulk download files
Team-based regulatory intelligence with collaboration	Commercial tool
One-time project with limited budget	FDA AccessData + openFDA API (both free)

Important: Regardless of which tools you use for research, your 510(k) submission must reference authoritative FDA data. Always verify critical information (K-numbers, decision dates, predicate details) against the official FDA database before citing them in a submission.

Integrating 510(k) Database Findings into Your Quality System

The 510(k) database is not just a research tool — its outputs should feed directly into your formal quality management system (QMS) processes. Regulatory intelligence from the database integrates with several key QMS elements.

Design Controls (21 CFR 820.30)

• Design input — Predicate analysis informs design requirements. The intended use, performance specifications, and testing approaches you find in predicate summaries should be captured as design inputs.
• Design verification and validation — The testing approaches described in predicate summaries establish a baseline for your own verification and validation plan. If every recent predicate in your product code includes biocompatibility testing per ISO 10993 and electrical safety testing per IEC 60601-1, your design validation plan should address these same areas.
• Design review — Predicate research findings should be presented at design reviews to confirm that the device under development is tracking toward a viable regulatory pathway.

Risk Management (ISO 14971)

• Hazard identification — MAUDE data cross-referenced with 510(k) clearances reveals known hazards for your device type. Adverse events reported for predicate devices and competitors should inform your hazard analysis.
• Risk evaluation — The frequency and severity of adverse events for devices in your product code help quantify risk levels.
• Risk control verification — Testing approaches in predicate summaries often reflect risk controls the FDA expects. Incorporate these into your risk control measures.

Regulatory Strategy Documentation

Document your 510(k) database research as part of your regulatory strategy file. This documentation should include:

• Search methodology (which fields, product codes, and date ranges you searched)
• Predicate candidate list with rationale for inclusion and exclusion
• Summary comparison tables for top predicate candidates
• Predicate selection decision with supporting evidence
• Cross-reference checks against MAUDE and Recalls databases
• Date of research (the database is updated weekly, so your research has a temporal context)

This documentation serves double duty: it supports your 510(k) submission and satisfies your QMS requirements for design input traceability.

Future Directions and Database Modernization

The FDA's 510(k) database and the broader 510(k) program are actively evolving. Understanding these trends helps you anticipate changes that may affect your regulatory strategy.

FDA Database Modernization

The FDA is investing in modernizing its IT infrastructure, including improved search interfaces for device databases. Recent additions of sub-category filters (PCCP, Clinical Trials, Combination Product) are part of a broader effort to provide more granular and useful data. Expect continued enhancements to search capability, data standardization, and cross-database linking.

Safety and Performance-Based Pathway

The FDA has been developing a Safety and Performance-Based Pathway that would allow certain 510(k) submissions to rely on performance criteria established through FDA guidance rather than comparison to a specific predicate device. This approach could reduce reliance on aging predicate chains and establish clearer, objective benchmarks for device performance. Several device categories are being evaluated for this pathway, and as it expands, it will create a new class of database records.

AI/ML and Predetermined Change Control Plans (PCCPs)

The rapid growth of AI/ML-enabled medical devices is fundamentally changing how the 510(k) database is used. Devices cleared with Predetermined Change Control Plans (PCCPs) can be modified within pre-approved parameters without requiring a new 510(k). This means the database record represents the initial clearance, but the device in the market may have evolved beyond what the original summary describes. When evaluating AI/ML-enabled devices as potential predicates, request the most current labeling and PCCP scope from the manufacturer.

Predictive Analytics

As the 510(k) dataset grows and computational tools mature, manufacturers and regulatory consultants are beginning to apply machine learning to predict clearance outcomes, estimate review timelines, and identify optimal predicate strategies based on historical patterns. While these approaches are still emerging, they represent a significant evolution from the manual, one-record-at-a-time research that has traditionally defined 510(k) database use.

Key Takeaways

• The FDA 510(k) database contains over 120,000 clearance records and is the primary tool for predicate research, competitive intelligence, and regulatory strategy in the US medical device market.
• Always start with product code searches — they are more reliable than device name searches because they bypass inconsistencies in how manufacturers name their devices.
• Search one field at a time, then narrow. Filling in multiple fields with the database's exact-match logic frequently returns zero results even when relevant records exist.
• 510(k) summaries are your most valuable resource — they reveal the predicate's intended use, testing approach, cited standards, and the basis for the SE determination. Prioritize predicates that have published summaries.
• The openFDA API enables programmatic access for trend analysis, automated monitoring, and large-scale data extraction. It is free and requires no authentication for basic use.
• Cross-reference multiple databases — the 510(k) database works best when combined with the Product Classification Database, MAUDE, the Recalls database, GUDID, and the De Novo database.
• The FDA's 2023 predicate selection guidance recommends using recently cleared predicates without design-related recalls, unmitigated safety issues, or use-related problems. Document your selection rationale.
• For complex predicate strategies, use the FDA's Pre-Submission (Q-Sub) program to get feedback before investing in testing.
• The database is updated weekly, but it does not include exempt devices, PMA devices, or pre-amendment devices that were never the subject of a 510(k). Never treat it as a complete inventory of all medical devices.
• Track your device category over time — evolving FDA expectations, new clearance patterns, and predicate eliminations directly affect your regulatory planning.
• Understand the SE framework — substantial equivalence requires same intended use plus same (or acceptably different) technological characteristics. Knowing this decision logic helps you evaluate predicate candidates and structure your own SE argument.
• Integrate database findings into your QMS — predicate research outputs feed directly into design controls (21 CFR 820.30), risk management (ISO 14971), and regulatory strategy documentation. Treat them as formal design inputs, not informal background research.
• Know the three 510(k) types — Traditional, Special, and Abbreviated submissions serve different purposes and have different review timelines. The type recorded in the database tells you about the nature of the clearance.
• Commercial search tools complement but do not replace the FDA's free databases. Use commercial platforms for efficiency and visualization, but always verify critical data against the authoritative FDA source.