Medtech 510(k) vs De Novo
510(k) vs De Novo: which FDA pathway should you file?
File a 510(k) when your device is substantially equivalent to a legally marketed predicate. Use the De Novo route when your device is novel and low-to-moderate risk but has no predicate — it would otherwise default to Class III. De Novo creates a new classification and is slower and more expensive, but it makes your device a future predicate.
| 510(k) | De Novo | |
|---|---|---|
| When to use | A moderate-risk device that is substantially equivalent to a device already on the market | A novel low-to-moderate-risk device with no suitable predicate (would otherwise be Class III by default) |
| Predicate device | Required — you must identify a legally marketed predicate and show substantial equivalence | Not required — the route exists precisely for devices without a predicate |
| What you ask FDA | To agree your device is as safe and effective as the predicate | To create a new classification and down-classify the device from the default Class III |
| Resulting class | Usually Class II (occasionally Class I) | Class I or Class II with special controls |
| Review-time goal (MDUFA V) | 90 FDA days for a MDUFA decision on a Traditional 510(k) | 150 FDA days for a De Novo decision |
| Standard user fee (FY2025) | Lower — roughly $24k (confirm current year) | Substantially higher — roughly $160k (confirm current year) |
| Outcome | Clearance to market | Grant + a new device type with special controls |
| Becomes a predicate? | No — you relied on an existing one | Yes — future devices can 510(k) to yours |
The decision turns on one question: does a suitable predicate exist?
If a legally marketed device is substantially equivalent to yours — same intended use, and the same or comparable technological characteristics — you file a 510(k) and demonstrate that equivalence. This is the common path for moderate-risk devices, and it is the faster, cheaper one: a 90-FDA-day MDUFA decision goal and a low-thousands user fee.
If your device is genuinely new — no predicate captures it — you cannot force a 510(k). Without a predicate, FDA’s default is to treat a novel device as Class III, which would push you toward a Premarket Approval (PMA). The De Novo route is the off-ramp: for a device whose risk is actually low to moderate, you ask FDA to create a new classification and place it in Class I or Class II with special controls. The trade is time and money — a 150-day review goal and a fee several times higher than a 510(k) — but the payoff is durable: your granted device becomes a predicate, so the next entrant has to 510(k) against you.
A practical sequence: search the FDA classification database and 510(k) database for a predicate first. If you find a clean one, file the 510(k). If you do not — or if a 510(k) comes back Not Substantially Equivalent — De Novo is usually the right pivot, provided the risk profile supports down-classification. If the risk is genuinely high, neither applies and you are in PMA territory.
Whichever route you take, the evidence burden underneath is similar: requirements traced to verification, predicate or special-controls comparison, and — for connected devices — the FDA Section 524B cybersecurity package. The pathway changes the wrapper and the timeline, not the need to assemble that evidence cleanly. For how long the 510(k) clock really runs once you file, see how long a 510(k) actually takes in 2026.
Sources
Last reviewed . Pathways and rules change — confirm against the cited primary source.