The Startup MDVisit Website →
← Back to blog

Avoid Months of Delay: De Novo vs 510(k) Strategy for U.S. SaMD Founders

September 15, 2026
Avoid Months of Delay: De Novo vs 510(k) Strategy for U.S. SaMD Founders

Use a 510(k) when you can point to a legally marketed predicate with the same intended use and show substantial equivalence to it. Use De Novo when no valid predicate exists, but the device carries low to moderate risk that general or special controls can reasonably manage. The 510(k) route asks you to prove sameness; the De Novo route asks you to prove safety and effectiveness from scratch, which usually means a heavier evidence package and a longer FDA review.

TL;DR:

  • A 510(k) requires a predicate device with similar intended use and technological features, whereas De Novo involves submitting safety and effectiveness data without a predicate.
  • De Novo grants create new product codes and set the standard for future devices, offering long-term market and regulatory advantages.
  • Evidence for a 510(k) is usually based on bench testing and verification, while De Novo often demands comprehensive nonclinical and clinical data to demonstrate safety.
  • Strategic decisions should consider whether general or special controls can manage safety risks, with De Novo suited for low to moderate risk devices without predicates.
The StartupMD
Plan Your SaMD Regulatory Strategy
The StartUp MD helps healthcare SaaS startups navigate growth challenges with tailored advisory and fractional consulting support.
Explore The StartUp MD

Table of Contents

De Novo vs 510(k): A Side-by-Side Comparison

The two pathways solve different problems, and mixing up which one applies to your device costs teams months. A 510(k) exists to confirm a new device is as safe and effective as something already on the market. De Novo exists for devices that have no such comparison point but still don't warrant the burden of a Premarket Approval (PMA).

The core distinction comes down to legal standard. A 510(k) requires substantial equivalence to a predicate under 21 CFR 807.81(a)(3). De Novo requires FDA to determine that general controls, or general controls plus special controls, provide reasonable assurance of safety and effectiveness for a device with no predicate, under 21 CFR 860.220.

  • Predicate required: 510(k) needs one; De Novo explicitly does not.
  • Evidence expectations: 510(k) typically leans on bench testing, performance data, and software verification and validation; clinical data enters only when the predicate comparison leaves a gap. De Novo almost always requires a fuller nonclinical package and often some clinical evidence to support the benefit-risk profile independently.
  • Review timelines: Between 2017 and 2023, the FDA processed roughly 2,929 510(k) submissions a year against just 26 De Novo requests annually, and mean review time for a 510(k) ran about 150 days, versus a 338-day mean and 309-day median for De Novo.
  • User fees: Standard 510(k) fees run lower than De Novo fees, and both carry substantial small-business discounts; check the current MDUFA fee schedule before budgeting, since rates adjust annually.
  • Strategic outcome: A 510(k) references an existing product code. A granted De Novo creates a brand-new one, and that device becomes the predicate everyone else in the category will cite for years.

That last point is easy to underweight when you're focused on getting cleared fast. It's the difference between joining an existing market category and defining one.

The 510(k) Pathway: What It Requires and How to Prepare

Substantial equivalence doesn't mean identical. FDA will accept a device with different technological characteristics than its predicate as long as those differences don't raise new questions of safety or effectiveness, and you can support that claim with data. A pulse oximeter with a redesigned sensor housing is a fine candidate. A diagnostic algorithm trained on a materially different population, making claims the predicate never made, is not.

Building a defensible 510(k) submission generally follows this sequence:

  1. Search the FDA 510(k) database for devices with matching intended use language, not just similar marketing names, and document why each candidate predicate does or doesn't fit.
  2. Select your predicate and write the comparison table covering intended use, technological characteristics, and performance specifications side by side.
  3. Build the evidence package. Bench testing and performance verification cover most hardware claims. Software-driven devices need documented verification and validation (V&V) tied to design controls under 21 CFR 820.30. Clinical data only enters the picture when bench testing can't answer the substantial-equivalence question on its own.
  4. Assemble the submission in eSTAR, FDA's structured electronic template, and run it against the acceptance-review checklist before you file, since a failed acceptance review restarts your clock.
  5. Anticipate the top reasons for a Not Substantially Equivalent (NSE) finding, most often a predicate with a mismatched intended use, insufficient performance data, or new technological characteristics that FDA believes raise unaddressed risk questions.

Pro Tip: Document your predicate rationale as if a second reviewer will read it cold. The single fastest way to trigger an FDA request for information is a comparison table that assumes context the reviewer doesn't have.

The De Novo Pathway: Eligibility, Content, and Strategic Weight

Illustration of staged De Novo pathway evaluation

De Novo exists for a specific situation: no legally marketed predicate exists, the device presents low to moderate risk, and general or special controls can manage that risk without the burden of a full PMA. Since FDASIA in 2012, sponsors can file directly without first collecting a 510(k) rejection, and most De Novo requests today are filed this way rather than as a fallback.

A complete request, built around the requirements in 21 CFR 860.220, needs to include:

  • A detailed device description and its precise intended use.
  • A benefit-risk analysis that stands on its own, since there's no predicate to lean on.
  • Supporting nonclinical data, and clinical data where the risk profile or novel mechanism demands it.
  • Proposed special controls, when you're recommending Class II designation instead of Class I.

The payoff for that heavier lift is real. A granted De Novo doesn't just clear your device. It creates a new product code and establishes the special controls that will govern the entire category going forward, and your device becomes the predicate every later 510(k) in that space has to reference. That's a durable regulatory advantage for the company that gets there first, but it comes at a cost: longer review, higher fees, and more uncertainty about what FDA will ultimately require before granting it.

How to Decide Between 510(k), De Novo, and PMA

Most teams overthink this decision or make it too late. A practical, stepwise approach keeps you from burning months on the wrong submission.

  1. Confirm the intended use and patient population precisely. A shift in population, indication, or clinical setting can eliminate predicates you assumed were solid.
  2. Run the predicate search and interpret the result honestly. If a predicate exists with matching intended use and comparable technology, 510(k) is very likely your path. If nothing matches, or every candidate has significant technological differences, De Novo becomes the default.
  3. Assess whether general or special controls can manage the device's risk. If yes, De Novo is viable. If the risk profile is high enough that controls alone can't provide reasonable assurance, you're looking at PMA territory instead.
  4. Do an evidence-gap analysis. Map what bench and performance data you already have against what each pathway would require, and be honest about whether clinical data is likely to be requested regardless of which route you pick.
  5. When the answer is unclear, or the stakes are high, schedule a Pre-Submission (Q-Sub). Bring specific, written questions about predicate selection, proposed testing, or classification, rather than a general "does this work" ask. A well-prepared Q-Sub is the cheapest insurance available against a rejected submission.

If a 510(k) comes back Not Substantially Equivalent, that's not a dead end. Filing a De Novo request is a recognized pivot, and some teams use the NSE as the signal to reframe the submission entirely rather than refiling a marginally stronger 510(k).

How Long Does FDA Review Actually Take?

Budget for reality, not the goal posted in the guidance document. MDUFA V sets a 150 FDA-day target for De Novo decisions, but observed performance runs well past that.

By the numbers: Mean 510(k) review time sits near 150 days, while De Novo decisions average 338 days, with a 309-day median, based on FDA data from 2017 through 2023.

That gap compounds with acceptance review, the initial FDA check confirming your submission is complete enough to enter substantive review. As of October 1, 2025, eSTAR is mandatory for De Novo submissions and applies to most 510(k)s as well, and a submission that fails the eSTAR acceptance checklist doesn't just wait. It restarts the clock. User fees scale with pathway too: standard De Novo fees run higher than standard 510(k) fees, with reduced rates for qualifying small businesses under the current MDUFA fee schedule.

Practical Tips for Software and AI-Enabled Devices

FDA reviewers of software-driven submissions focus on a specific set of documentation, and gaps here generate more requests for information than almost anything else in a SaMD file.

  • Document algorithm performance and the clinical rationale behind it, not just the accuracy metrics, and be explicit about training and test data provenance.
  • Tie your design-control records under 21 CFR 820.30 directly to the sections of your submission that reference them, rather than leaving reviewers to infer the connection.
  • Treat IEC 62304 software lifecycle evidence as premarket documentation, not an internal engineering artifact.
  • Address cybersecurity and off-the-shelf software components explicitly, since silence on either invites a follow-up request.

Pro Tip: Run a focused Q-Sub on your algorithm's clinical rationale before you finalize the submission. Reviewers who see clear data provenance and benefit-risk framing early tend to send fewer RFIs later.

Where to Verify These Requirements Directly

Why Speed Isn't Always the Right Metric

Founders often ask which pathway is faster, but that's the wrong first question. The right one is which pathway matches your actual risk profile and evidence, because filing the wrong one costs more time than either route does on its own.

De Novo is worth the longer timeline when a grant hands you a new product code and a durable predicate advantage that later entrants have to build around. A staged product release, where a lower-risk version clears through 510(k) first while you gather evidence for a more capable follow-on, often beats waiting for one perfect De Novo submission. What consistently separates a clean review from a stalled one isn't the pathway chosen. It's whether the clinical and regulatory strategy was built before the submission, not patched together after FDA's first request for information.

— Paul Bergeron MD, MBA

Get Pathway Strategy Right Before You File, Not After

Choosing wrong between 510(k) and De Novo doesn't just cost you FDA's review clock. It costs your fundraising timeline, your go-to-market plan, and often a full resubmission cycle. Specialized consultancies work with healthcare SaaS and SaMD founders as fractional clinical and regulatory advisory partners, building pathway strategy, evidence plans, and Pre-Submission questions before they become expensive corrections after an RFI.

The StartupMD

That means regulatory strategy sessions to map your device against predicate options, clinical evidence planning proportional to your actual risk profile, and Q-Sub preparation designed to surface FDA's real concerns before you file. Founders who bring in that structure early tend to submit better-aligned packages and face fewer RFIs, which keeps timelines predictable instead of guesswork. If you're weighing pathway choice against your funding runway and go-to-market plan, start with The StartupMD's Healthcare SaaS Revenue Model Evaluation to see how regulatory strategy and commercial timing fit together, and book a conversation about your specific device before you commit to a submission strategy.

This article is general information, not a substitute for advice from a qualified doctor. Consult a qualified healthcare professional about your own circumstances before acting on anything here.

Sources

FAQ

How much does a De Novo request cost?

FDA publishes standard and small-business De Novo user fees each fiscal year under the MDUFA fee schedule, and De Novo fees run notably higher than standard 510(k) fees. Check FDA's current fee schedule directly, since rates adjust annually and vary by applicant size.

Is 510(k) clearance the same as FDA approval?

No. A 510(k) results in "clearance," meaning FDA found the device substantially equivalent to a predicate, while "approval" is the term reserved for PMA decisions, which require independent proof of safety and effectiveness. A granted De Novo results in "grant," not approval or clearance, though the practical market effect is similar to clearance.

Is a 510(k) the same as a PMA?

No. A 510(k) relies on substantial equivalence to an existing predicate device, while a PMA requires independent clinical evidence of safety and effectiveness for higher-risk Class III devices with no equivalence pathway available.

What is De Novo for FDA?

De Novo is the risk-based classification pathway for novel devices with no legally marketed predicate. A grant establishes a new product code, and that device becomes the predicate other companies can cite in future 510(k) submissions.