.png){kind=logo}
If your medical device is genuinely novel, the default under the Federal Food, Drug, and Cosmetic Act is not kind to you. A device type that has never been classified is automatically Class III under section 513(f)(1), the same statutory category as the highest-risk devices, regardless of how modest the actual risk profile is. The De Novo pathway exists to fix that mismatch: it gives low-to-moderate-risk, first-of-a-kind devices a route to market with a risk-based classification instead of a premarket approval (PMA).
The pathway has changed meaningfully in recent years. The process is now codified in regulation, electronic submission is mandatory, and review timelines carry their own MDUFA performance goals. Below is a current crash course: how De Novo works in 2026, what it costs, what FDA expects in a request, and how to decide whether it is the right route for your device.
What the De Novo pathway is
De Novo, formally the De Novo classification process under section 513(f)(2) of the FD&C Act, lets a sponsor ask FDA to classify a novel device into Class I or Class II based on its actual risks rather than leaving it in Class III by default. When FDA grants a De Novo request, three things happen at once:
- Your device is classified into Class I or Class II and can be marketed immediately.
- FDA establishes a new classification regulation for the device type, including special controls for Class II devices that define the testing, labeling, and design expectations for every future device of that type.
- Your device becomes eligible to serve as a predicate: later entrants (and your own next-generation products) can reach market through the far less expensive 510(k) pathway by claiming substantial equivalence to it.
If FDA declines the request, the device remains Class III. The realistic options at that point are a PMA or rethinking the indications for use and evidence package and trying again.
Two ways into De Novo
After an NSE determination
If you submitted a 510(k) and FDA found the device not substantially equivalent (NSE), for example because there is no adequate predicate, the device has a new intended use, or different technological characteristics raise different questions of safety and effectiveness, you may submit a De Novo request asking FDA to classify the device based on risk.
Direct De Novo
If you already know there is no legally marketed device on which to base a substantial equivalence claim, you do not need to file a 510(k) first. You can submit a De Novo request directly. Since FDASIA created this option in 2012, the direct route has become the common one.
Choosing between a defensible 510(k) and a direct De Novo is one of the most consequential early strategy calls for a novel device. Guessing wrong can cost a full review cycle: a 510(k) that ends in an NSE determination means months of review time spent to arrive where a classification analysis would have pointed you on day one. A Pre-Submission is the standard way to pressure-test the pathway question with FDA before committing.
The process is now codified: 21 CFR Part 860, Subpart D
Since January 2022, the De Novo process has been codified at 21 CFR 860.200 through 860.260. The final rule replaced guidance-only expectations with binding requirements:
- Content requirements (21 CFR 860.220). A De Novo request must include, among other elements, a device description, proposed indications for use, a classification summary explaining why no predicate exists, the probable benefits and probable risks to health of the device type, proposed mitigations (the seed of the special controls), and the performance data supporting them.
- Acceptance review. FDA screens each request for completeness before substantive review begins. An incomplete request is refused before the review clock ever starts.
- Facility inspections. The rule permits FDA to inspect facilities where the device is designed, manufactured, or tested as part of its De Novo review, before granting the request.
Electronic submission is mandatory. Since October 1, 2025, De Novo classification requests must be prepared using FDA's eSTAR template and submitted electronically through the CDRH Customer Collaboration Portal. Paper submissions and eCopies are no longer accepted for De Novos. If your team has only built 510(k) eSTARs, note that the De Novo eSTAR carries additional sections aligned to 860.220, including the classification summary and proposed special controls.
Timelines and fees in FY2026
Under MDUFA V, FDA's performance goal is a MDUFA decision within 150 FDA review days, applicable to 70 percent of De Novo requests received in FY2026. FDA days pause while the agency waits on your responses to deficiency letters, so total calendar time is typically well beyond 150 days. A year from submission to grant is a realistic planning assumption, and complex requests can run longer.
The FY2026 standard user fee for a De Novo request is $173,782, with a small business rate of $43,446 (25 percent of the standard fee). For comparison, an FY2026 510(k) costs $26,067 standard and $6,517 small business. The fee gap is one more reason the classification question deserves rigor early, and if you may qualify as a small business (gross receipts of $100 million or less), certify before you submit: the decision cannot be applied retroactively to a fee already paid.
What changes under MDUFA VI
MDUFA V expires at the end of FY2027, and the draft MDUFA VI commitment letter (covering FY2028 through FY2032) points to a faster, more structured De Novo program. The headline change: the decision goal tightens from 70 percent to 90 percent of De Novo requests receiving a MDUFA decision within 150 FDA days. The draft also adds structure to the review itself. For De Novos received on or after October 1, 2027, FDA and the requester will hold an Introduction Meeting within the first 30 FDA days of review. Deficiency letters must reflect a complete review, include all deficiencies with a statement of their basis, and pass supervisory review before issuance, with any follow-up deficiencies limited to issues raised by your response. If FDA issues a not-grantable letter, it intends to reach a final grant-or-decline determination within 75 calendar days of your response (and no later than 90), and if no decision has been reached by 180 FDA days, FDA will walk through every outstanding issue with you, with action items and estimated completion dates on both sides. FDA also plans to pursue a policy setting the timeframe for sponsors to respond to deficiency and not-grantable letters, on the order of 360 calendar days.
The caveats: this is a draft commitment letter, and MDUFA VI still requires congressional reauthorization before the FY2028 cycle begins, so the numbers can move and FY2028 fee amounts are not yet set. Directionally, though, the agreement describes a De Novo program with earlier engagement, more predictable deficiency practice, and tighter timelines. You can download the draft commitment letter from our resources library: Draft MDUFA VI Commitment Letter (FY 2028-2032).
What a strong De Novo request looks like
A De Novo flips the logic of a 510(k). Instead of comparing your device to a predicate, you are building the case that the risks of a brand-new device type can be identified and controlled. FDA's review centers on four things:
- A complete risk picture. Identify the probable risks to health for the device type as a whole, not just your implementation of it. FDA must be able to see that the risks are identifiable and bounded.
- Proposed special controls. Map each identified risk to a specific mitigation: performance testing, labeling, design characteristics, or usability work. Granted De Novos in adjacent spaces are the best template; their classification orders and decision summaries show exactly the form FDA expects special controls to take.
- Performance data tied to the intended use. Acceptance criteria need clinical justification for your specific use context and anatomy of interest, and testing should reflect the intended use population, users, and use environment. Criteria borrowed from a different clinical context invite deficiency letters.
- Benefit-risk. The request must support a finding that, with the special controls applied, the probable benefits outweigh the probable risks and there is reasonable assurance of safety and effectiveness.
Do the Pre-Sub. For nearly every De Novo candidate, a Pre-Submission to align on the testing plan, the proposed special controls, and clinical evidence expectations is worth the calendar time. It is far cheaper to learn FDA's view of your acceptance criteria before the studies are run than after.
What this means for developers
Ask the classification question early. Before committing to a predicate strategy, interrogate whether your intended use truly matches it. A new intended use or a new clinical workflow is exactly what turns a planned 510(k) into an NSE letter.
Write the special controls, do not wait for them. Companies who arrive with a credible risk-to-mitigation map shape the conversation. Companies who wait for FDA to define the controls inherit whatever the review team writes.
Budget realistically. Plan for the FY2026 fee, a year of review, at least one substantive deficiency cycle, and the Pre-Submission in front of all of it.
Think past your own authorization. A granted De Novo defines the classification regulation your competitors will use to follow you through 510(k)s, and the predicate your own adjacent and next-generation devices can build on. Draft the indications for use and special controls with the whole product roadmap in mind, not just the first device.
The bigger picture
De Novo has moved from a niche workaround to a central pathway for novel technology, particularly AI-enabled devices and software whose new intended uses outrun the existing classification regulations. FDA has also signaled, through NSE determinations and guidance, that it reads intended use narrowly when new clinical workflows introduce new types of risk, which pushes more first-movers toward De Novo whether they planned for it or not. For a deeper look at how to engage FDA before you commit to a pathway, see our post on FDA Pre-Submissions.
How Cosm Can Help
Cosm supports medical device, SaMD, and AI/ML companies through exactly these decisions: pathway strategy, Pre-Submissions, De Novo requests, special controls development, and the performance testing plans behind them. If you are weighing a 510(k) against a De Novo, or responding to an NSE determination, reach out at info@cosmhq.com or visit cosmhq.com.
Disclaimer - https://www.cosmhq.com/disclaimer
