Predicate Devices: How to Find and Choose One for Your 510(k) Submission

2026-07-30 · Caduvo Team

A practical guide for regulatory affairs teams: what a predicate device is, how to search the FDA 510(k) database, and how to rank candidates to minimize testing burden and review risk.

Your 510(k) submission can live or die by the predicate device you select. Get it wrong and you’re looking at avoidable testing, extra review questions, or a refusal to accept. Get it right and the path to clearance shortens dramatically. This walkthrough gives regulatory affairs teams a concrete, repeatable method for finding and ranking predicate candidates — and shows why the choice shapes your entire submission strategy.

What Is a Predicate Device and How Does Substantial Equivalence Work?

A predicate device is a legally marketed device that you cite in your 510(k) submission to demonstrate that your device is substantially equivalent. That’s the core of a 510(k): you’re not proving independent safety and effectiveness from scratch. Instead, you’re comparing your device to one that’s already on the U.S. market and was never removed for safety reasons.

Substantial equivalence means two things line up: intended use and either technological characteristics or a demonstration that any differences don’t raise new questions of safety and effectiveness. The intended use must be exactly the same — same clinical condition, same patient population, same treatment approach. If your device treats the same condition but uses a fundamentally different mechanism of action, you can’t just claim substantial equivalence by paper comparison. You’ll need to show, through testing, that the differences don’t matter.

A strong predicate has an intended use that closely mirrors yours and technological characteristics that are similar enough that the testing burden stays low. What is a predicate device in practice? It’s the benchmark against which FDA measures your device, so the closer the match, the fewer questions reviewers will raise.

Searching FDA Databases for Predicate Candidates

Start with the FDA 510(k) Premarket Notification database. You don’t need special access. The search workflow looks like this:

Find your likely product code using the FDA Product Classification database. Identify the regulation number under which similar devices are classified. FDA medical device classes and corresponding product codes narrow the field quickly. For example, a nonabsorbable suture might map to product code GAT under 21 CFR 878.4490.
Search the 510(k) database by product code. Enter the three-letter code in the “Product Code” field. Filter by clearance date — start with the last 5 to 7 years. Devices cleared more than a decade ago may rely on standards or test methods that are now obsolete, triggering modern expectations even if the older predicate sailed through.
Scan device names and applicants. Open the 510(k) summaries (K numbers) for devices that sound like yours. Look at the intended use statement and the indications for use — those are often verbatim in the summary. If they differ from yours, you’ll have to justify that difference.
Review the “substantial equivalence” discussion in each summary. That section often names the predicate that device used. Following the chain can reveal whether your prime candidate has strong lineage or is itself a stretch.

This manual search works, but when you’re comparing a dozen candidates across multiple product codes or months of review timelines, the work compounds. Caduvo’s platform automates much of this cross-referencing, but the logic you apply manually is the same.

Ranking Predicate Candidates: What Makes a Strong Choice

Once you have a list, rank candidates on these criteria:

Intended use match (non-negotiable). Same indication, same anatomical site, same patient population, same duration of use. Partial match means you’ll need performance data to bridge the gap.
Technological similarity. If your device uses the same materials, energy source, and operating principle, you’re in the best position. A different sensor type or coating can trigger biocompatibility or performance tests you could have avoided.
Recency of clearance. Devices cleared within the last 5 years reflect current FDA expectations. Older predicates might have been cleared under less rigorous standards, so FDA may request testing that wasn’t required before.
Clearance pathway and testing burden. Read the 510(k) summary to see what data was submitted. If the predicate required a clinical study, you’ll almost certainly need one too — unless you can argue that your differences eliminate the need. If the predicate cleared with bench testing only and you match it closely, that’s a strong signal.
Manufacturer reputation and prior interactions. A predicate from a well-known manufacturer with a clean regulatory history reduces unknowns. A device that’s been subject to recalls or adverse events will draw more scrutiny.

Rank multiple candidates. Your first choice may look strong, but a backup with slightly older clearance but a near-identical intended use can be a safer bet if the primary hits a snag during pre-submission.

Risks of a Weak or Split Predicate Strategy

Choosing a weak predicate — one with a different intended use or dissimilar technology — forces you to generate more data. You might need a comparative bench study, animal data, or a clinical trial to prove that differences don’t affect safety and effectiveness. That adds months and tens or hundreds of thousands of dollars to your timeline.

A split predicate strategy uses multiple predicates to cover different aspects of your device (e.g., one for software, another for the hardware platform). This is sometimes necessary for combination products, but it’s riskier for standard devices. FDA reviewers will ask why no single device is comparable. If the split seems contrived to avoid testing, expect a request for additional information or a refusal to accept. Only use a split predicate when you’ve exhausted every search for a single match and you’ve documented that no single device fills the role.

Old predicates carry another danger: design controls and consensus standards have evolved. A device cleared in 2008 may not reflect current biocompatibility or software standards. Even if you match it perfectly, FDA can ask you to meet today’s requirements, erasing the advantage you sought.

Actionable Takeaways

Begin predicate research at the concept stage — not after design freeze. A late find can force a design change or a testing sprint.
Map your intended use statement in writing before you search. Vague wording leads to vague matches.
Use the product code to filter systematically; don’t rely on keyword searches alone.
Rank candidates by intended use match, technological similarity, and recency, in that order.
Read the 510(k) summary for every top candidate, paying special attention to the indications for use and the data submitted.
Avoid a split predicate unless you have documentation showing why no single predicate exists.
When in doubt, request a pre-submission meeting and propose your top candidate. FDA will often tell you if it’s appropriate before you invest in full testing.