A granted De Novo becomes a predicate.

Every device not in commercial distribution before May 28, 1976 starts in Class III by operation of law — not because anyone assessed its risk, but because that is the statute’s default when there is nothing to compare the device to. Most new devices escape Class III through a 510(k), by matching something already on the market. The De Novo pathway exists for the device that cannot: genuinely low-to-moderate risk, and no legally marketed predicate.

The route used to run through rejection

Before 2012, reaching De Novo meant manufacturing your own rejection first. A sponsor filed a 510(k) it expected to lose, waited for FDA to issue a Not Substantially Equivalent determination, and only then petitioned to have the device reclassified out of Class III. The statute required the failure as a precondition — a procedural detour with nothing to do with the device itself, and months added to the calendar before the real review could start.

Section 607 of the FDA Safety and Innovation Act (FDASIA) — Public Law 112-144, enacted July 9, 2012 — amended Section 513(f)(2) of the FD&C Act (21 U.S.C. 360c(f)(2)) to add a second door: a sponsor who has already determined no legally marketed predicate exists can request De Novo classification directly, without ever filing the 510(k) built to fail. The NSE-first route still exists. It is no longer the only one.

What a grant actually creates

A De Novo grant classifies the device into Class I or Class II — with special controls attached, for Class II — and FDA’s own guidance says plainly what happens next: the granted device can be used as a predicate for future 510(k) submissions of devices of the same type. The device that had nothing to point to becomes, the moment the order issues, the thing later devices point to.

A pending De Novo has no predicate value. A granted one is somebody else’s.

That timing detail matters more than it sounds. “No predicate exists” describes a moment, not a permanent fact about a device type — and the moment ends the day FDA grants the request. A competitor citing a De Novo device as its predicate is telling you something concrete and checkable: that grant is final, on the record, sitting in FDA’s classification database with its own regulation number. A citation to a still-pending De Novo request is not a citation to anything yet — it is a citation to an application.

The clock is negotiated, not written into the statute

Section 513(f)(2) sets no deadline for FDA’s decision. The number people quote — 150 days — comes from the MDUFA V commitment letter covering fiscal years 2023 through 2027: a negotiated performance goal, starting at 70 percent of De Novo requests decided within 150 FDA days in FY2023, stepping higher only if fee revenue and prior-year performance both hold. It is a target FDA agreed to chase, renegotiated every five years alongside the user-fee reauthorization — not a right a sponsor can enforce the way a statutory deadline would be.

The practice

Before treating any device as a usable predicate, check which side of the grant it sits on. FDA’s classification database — the same one that derives class, regulation, pathway, and review panel from a product code on FDA Radar — only reflects a device once FDA has actually classified it. A De Novo request that has not yet been granted will not appear there as anything citable, no matter how confident the sponsor’s own marketing sounds. This is regulatory intelligence and method, never regulatory advice about whether a specific device or comparison supports your own submission; that judgment belongs to the regulatory professional holding the file.