FDA 510(k) Predicate Device Search: Step-by-Step Database Workflow

Quick answer

To find an FDA 510(k) predicate device, first determine your device's product code and classification, then search FDA's releasable 510(k) database by product code, device name, applicant, or 510(k) number. A valid predicate must be a legally marketed U.S. device that supports a substantial equivalence (SE) comparison on intended use and technological characteristics. The predicate is not simply a commercially similar product — it is the legally marketed reference device named in your 510(k) comparison under 21 CFR 807.87.

Who this applies to

This workflow applies to U.S. medical device manufacturers and U.S. agents preparing a Premarket Notification 510(k) for a device that is not PMA-exempt by class and not eligible for a simpler pathway. It is most useful when:

• You know your device class but have not yet locked a predicate strategy
• Your team needs to compare multiple candidate predicates before pre-submission
• You are evaluating whether differences in materials, software, or energy delivery require additional performance data

It does not replace FDA's full SE analysis. Database search results show what was cleared — not automatic acceptance of your predicate choice.

What counts as a predicate device

FDA defines substantial equivalence relative to a legally marketed predicate device. Per FDA predicate guidance and the 510(k) overview, a predicate may include:

• A device recently cleared through 510(k)
• A preamendments device legally on the U.S. market before May 28, 1976
• A device that was Class III (PMA) and later downclassified to Class I or II
• A 510(k)-exempt device, in limited circumstances described in FDA guidance

Legally marketed means the predicate cannot be a device that is or was in violation of the Federal Food, Drug, and Cosmetic Act. FDA also notes you may, in some cases, reference a predicate no longer actively marketed in the U.S. if it remains legally marketed — verify carefully against current FDA guidance before relying on a discontinued predicate.

Primary predicate vs reference devices

FDA recommends identifying a primary predicate — the device most similar to your subject device with respect to indications for use and technological characteristics. You may identify more than one predicate in certain circumstances, but split predicates (combining features from unrelated predicates to justify SE) are inconsistent with the 510(k) standard.

You may also identify reference devices to support scientific methodology or standard reference values without claiming them as predicates.

Substantial equivalence: what you are actually comparing

FDA's 510(k) framework requires demonstrating SE to a predicate. At a high level, a device is substantially equivalent if:

1. It has the same intended use as the predicate; and
2. It has the same technological characteristics as the predicate; or
3. It has the same intended use and different technological characteristics that do not raise different questions of safety and effectiveness, and submitted information demonstrates the device is as safe and effective as the legally marketed device.

SE is not identity. FDA evaluates differences in design, materials, energy, software, biocompatibility, sterility, labeling, and performance testing as part of the comparison.

Decision flow before you search

Step-by-step predicate search workflow

Step 1 — Confirm classification and product code

Use FDA Medical Device Classification resources and the Product Classification Database to identify:

• Device class (I, II, or III)
• Regulation number (e.g. 21 CFR 870.xxxx)
• Panel (e.g. CV, GU, HO)
• Three-letter product code (e.g. DQX, FRN)

The product code is often the highest-yield search key because device trade names vary across manufacturers.

Official resource: FDA Medical Device Databases — https://www.fda.gov/medical-devices/device-advice-comprehensive-regulatory-assistance/medical-device-databases

Step 2 — Search the releasable 510(k) database

Use FDA's Search the Releasable 510(k) Database to query by:

• Product code (recommended first pass)
• 510(k) number (when you already have a lead clearance)
• Device name
• Applicant name
• Panel

For each candidate record, download or review:

• 510(k) summary (if available)
• Indications for use statement
• Technological characteristics described in the clearance
• Any special controls or labeling conditions

Official resource: https://www.fda.gov/medical-devices/510k-clearances/search-releasable-510k-database

Step 3 — Build a predicate shortlist matrix

Document each candidate in a comparison table before selecting a primary predicate:

Reject candidates where intended use diverges — even if the hardware looks similar.

Step 4 — Read the predicate 510(k) summary like FDA does

For your top one to three candidates, read the cleared device's public summary and ask:

• What indications for use were cleared?
• What technological characteristics did FDA compare?
• Did FDA require clinical data for differences?
• Are there labeling limitations you would inherit?

If your device adds a new user population, new anatomical site, new energy level, or new software algorithm with different clinical impact, treat that as a different question of safety and effectiveness until proven otherwise.

Step 5 — Lock primary predicate and draft comparison outline

Your 510(k) comparison (per 21 CFR 807.87) should identify:

• Primary predicate 510(k) number(s)
• Side-by-side intended use comparison
• Side-by-side technological characteristics
• Rationale for why differences do or do not raise new questions
• Performance testing mapped to each difference

Consider a Pre-Submission (Q-Sub) meeting if predicate selection is uncertain.

Key documents to prepare alongside predicate research

Checklist: predicate search and selection

• Product code confirmed via FDA classification database
• Releasable 510(k) database searched by product code first
• At least three candidate predicates reviewed
• Intended use compared verbatim, not paraphrased loosely
• Technological differences listed explicitly (materials, design, energy, software)
• Split-predicate strategy avoided
• Primary predicate identified with 510(k) number
• Predicate legally marketed status confirmed
• Performance testing plan mapped to each difference
• Predicate selection rationale documented for design history file

Common pitfalls

1. Searching by brand name only — competitors often use different trade names under the same product code.
2. Choosing the newest clearance — recent is helpful for review habits, but SE depends on similarity, not recency.
3. Ignoring indications for use — hardware similarity with different clinical claims fails SE.
4. Split predicates — mixing intended use from one device and technology from another.
5. Assuming database presence equals acceptability — FDA still judges your comparison and data.
6. Using non-U.S. CE-marked devices as predicates — foreign approvals are not U.S. predicates.
7. Skipping reference device strategy — useful for test methods without naming extra predicates.

FAQ

How many predicate devices can I use in one 510(k)?

FDA allows more than one predicate in certain circumstances, but you should designate a primary predicate most similar on intended use and technology. Multiple predicates do not bypass the SE standard.

Can I use a device that is no longer sold in the U.S.?

FDA guidance states that, in some cases, a predicate no longer marketed may still be legally marketed. Confirm against current FDA predicate guidance and document your rationale carefully.

What is the difference between a predicate and a reference device?

A predicate is the legally marketed device to which you claim substantial equivalence. A reference device supports test methodology or standard values and is not a substitute predicate.

What FDA database should I search first for predicates?

Start with the Product Classification Database to find your product code, then search the releasable 510(k) database by that code.

Do I need clinical data if my device differs from the predicate?

Not always. If technological differences do not raise new questions of safety and effectiveness, bench and other non-clinical data may suffice. If new questions arise, clinical or additional performance data may be required.

What are split predicates and why are they problematic?

Split predicates combine features from multiple devices in ways that do not reflect a single legally marketed device's intended use and technology. FDA considers this inconsistent with the 510(k) standard.

Can a 510(k)-exempt device be a predicate?

FDA guidance lists limited circumstances where exempt devices may ...