How do I find similar medical devices already cleared by FDA?
To find similar medical devices already cleared by FDA, four public sources are the starting point. As of 2026-05-22, those are: (1) the FDA 510(k) database, searchable by product code, applicant, device name, and decision date; (2) the FDA PMA database, for Class III approved devices; (3) FDA Product Classification, which maps a device description to a candidate FDA product code, CFR citation, panel, and class; and (4) openFDA, the programmatic API. This workflow walks each source and shows how to compile a candidate predicate shortlist with K-numbers, applicants, decision dates, and product codes. Predicate selection — choosing which similar device to cite in a 510(k) submission — is a strategic decision made by the manufacturer's RA/QA team or qualified consultant. This workflow surfaces the shortlist; it does not select the predicate, does not determine substantial equivalence, and does not produce a regulatory determination of any kind.
For founders, manufacturers, RA/QA teams, and regulatory consultants. For RA/QA or consultant review.
This is not regulatory, legal, or compliance advice. TrueMedDevice does not determine FDA classification, submission pathway, predicate status, substantial equivalence, compliance, safety, or effectiveness. We prepare a source-backed evidence pack from public FDA sources so qualified RA/QA professionals or regulatory consultants can review and decide.
The four public FDA sources at a glance
Each source answers a different research question. Most candidate predicate shortlists draw on at least three of the four.
- FDA 510(k) Database
Search by product code, applicant, device name, decision date.
- FDA Premarket Approval (PMA) Database
Approved Class III devices and their supplements.
- FDA Product Classification
Map device description to candidate product code, CFR citation, panel, and class.
- openFDA
Programmatic API access to 510(k), PMA, MAUDE, recalls, and more.
Step 1 — Candidate product-code clues
FDA Product Classification is the entry point. Search by device description (e.g., "infusion pump," "administration set," "catheter," "automated external defibrillator") and the database returns three-letter product codes paired with the relevant 21 CFR regulation citation, the device class (I, II, III), and the review panel.
More than one candidate often applies. Note them all as candidate clues — selection of the final product code is FDA's determination when a submission is accepted. The candidate-clue framing is the right discipline at this stage.
What this surfaces: candidate product codes, candidate device-class indications, and the CFR section that governs each candidate.
Step 2 — FDA 510(k) database search
With a candidate product code in hand, the 510(k) database returns every device cleared under that code. The search supports applicant, device name, 510(k) number (K-number), and decision date filters. Each cleared entry links to a 510(k) Summary or 510(k) Statement, and sometimes a Decision Summary.
Read the Summaries to understand how each cleared device describes its intended use, indications, technological characteristics, performance, and labelling. The structured similar-product comparison you build off these documents is the substance of a candidate predicate shortlist.
What this surfaces: candidate predicate devices with K-numbers, applicants, decision dates, and the textual basis for similar-product comparison.
Step 3 — Map the similar-device landscape
A predicate shortlist is rarely a single device. Useful shortlists include direct similar products (close on intended use and technological characteristics) and indirect similar products (relevant on one axis but distinct on another). The point is to understand the category, not to pre-commit to one candidate.
A similar-product comparison table — built from public 510(k) Summaries, public manufacturer product pages, IFUs, and labels — is the cleanest output. Each row in the table is a source-cited data point a reviewer can re-walk.
What this surfaces: a structured similar-product comparison with direct and indirect candidates, source-cited for every row.
Candidate predicate vs. final predicate — the boundary
A research workflow surfaces candidate predicates. The final predicate is a regulatory selection — made by the manufacturer's RA/QA team in consultation with a qualified consultant and ultimately validated by FDA when it accepts a 510(k) submission on substantial-equivalence grounds.
That separation matters because substantial equivalence is FDA's determination, never the research output. The candidate shortlist is an input to that determination — never a substitute for it.
What this surfaces: clear separation between research output (candidate clues, candidate predicate shortlist, similar-product comparison) and regulatory determination (final predicate selection, substantial-equivalence finding).
Step 4 — Why recall, MAUDE, and warning-letter checks matter
Each candidate predicate carries a safety history. FDA MAUDE (Manufacturer and User Facility Device Experience) records voluntary adverse event reports; FDA Recalls records voluntary and involuntary corrective actions classified by FDA as Class I, II, or III; FDA Warning Letters records enforcement actions; FDA Safety Communications records public advisories.
A candidate predicate with significant safety history is not disqualified — but the information shapes how a 510(k) submission argues similar-product comparison and what testing the manufacturer plans. MAUDE counts are not incidence rates; reports are voluntary and unverified.
Absence claims must be scoped and dated. "No warning letters found in the FDA Warning Letter database for manufacturer X as of 2026-05-22" is a defensible statement. "Manufacturer X has no compliance issues" is not.
How this prepares a better consultant or RA/QA conversation
A founder who arrives at a regulatory consultant call without a candidate predicate shortlist typically spends the first paid hour explaining the device and watching the consultant build the shortlist they could have built themselves. Bringing a source-backed evidence base — candidate product-code clues, candidate predicate shortlist, similar-product comparison, safety-history scan — shortens discovery and shifts the conversation toward judgment: which predicate to select, what testing strategy follows, whether a Pre-Submission (Q-Sub) meeting with FDA is warranted.
The candidate predicate shortlist is not the answer; the consultant's judgment is. The shortlist is what makes that judgment cheaper and faster.
Frequently asked questions
What is a predicate device?
A predicate device is a U.S. medical device that has already been cleared by FDA via the 510(k) pathway, and that a manufacturer cites in a new 510(k) submission to argue that their device is similar enough to be cleared on the same basis. Predicate selection is the manufacturer's strategic decision, made with RA/QA and consultant input. This workflow surfaces candidate predicate devices from public FDA data — it does not select the predicate for you and does not determine substantial equivalence.
What's the difference between FDA-cleared and FDA-approved?
"FDA-cleared" refers to devices that have been authorized via the 510(k) pathway based on substantial equivalence to a predicate device. "FDA-approved" is reserved for Class III devices authorized via the Premarket Approval (PMA) pathway. Most marketed medical devices are cleared, not approved. Using "FDA-approved" to describe a 510(k) device is a factual error.
Do I need a U.S. agent if I'm a foreign manufacturer?
FDA requires every foreign establishment that imports or offers for import medical devices into the U.S. to designate a U.S. agent. The U.S. agent is the formal communication channel between FDA and the foreign establishment for routine and emergency communications. U.S. agent selection is the manufacturer's decision; this research workflow surfaces the obligation context from public FDA sources for RA/QA or consultant review.
How do I search the FDA 510(k) database?
The FDA 510(k) database lives at accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm. Search by applicant, device name, 510(k) number, product code, decision date, or several other fields. Each cleared entry links to a 510(k) Summary or 510(k) Statement, and sometimes a Decision Summary, that you can read to compare your candidate device against the cleared one. The openFDA API at api.fda.gov also provides structured access to the same records.
What evidence should I bring to a regulatory consultant?
Before a paid consultant call, gather: (1) a written intended-use statement, (2) candidate FDA product-code clues with sources, (3) candidate submission-pathway clues with rationale, (4) a candidate predicate shortlist with K-numbers, (5) a safety-history scan of similar devices (MAUDE, recalls, warning letters), and (6) a forms / fees / typical review-timeline reference. The Product Evidence Pack delivers each of these as a source-cited artifact. Classification, pathway, predicate selection, and submission strategy decisions remain with your RA/QA team or qualified consultant.
Want this workflow run on your product?
The U.S. / FDA Market-Entry Product Evidence Pack delivers the workflow above — candidate product-code clues, candidate predicate shortlist, similar-product comparison, safety-history scan, source ledger — as a source-cited artifact for qualified RA/QA or consultant review.
See the U.S. / FDA Market-Entry Product Evidence PackThis is not regulatory, legal, or compliance advice. TrueMedDevice does not determine FDA classification, submission pathway, predicate status, substantial equivalence, compliance, safety, or effectiveness. We prepare a source-backed evidence pack from public FDA sources so qualified RA/QA professionals or regulatory consultants can review and decide.