Quick answer: Whether you need an FDA 510(k) consultant depends less on your budget than on three things: how many submissions you'll file, how well-established your predicate device is, and whether anyone on your team has done this before. A manufacturer with a clear predicate, in-house regulatory affairs depth, and no urgent deadline can genuinely run a 510(k) in-house — plenty do it well. The risk rises sharply for a first submission, an ambiguous predicate, or a hard deadline, because a wrong call on predicate selection or substantial equivalence doesn't just cost a consultant's fee — it costs a full additional review cycle, and FDA's review clock pauses, not stops, while you fix it. Most manufacturers land on a hybrid: strategy and predicate judgment from someone who has done this before, documentation and submission assembly kept in-house.
The real question isn't "should I outsource"
It's: how many times am I going to do this, and how much ambiguity am I starting with?
Framed as a binary, the decision invites a bad answer either way. "Always outsource" ignores that a repeat filer with a stable product line and an in-house RA lead builds real capability over successive submissions — capability that's wasted if every filing goes to an outside firm. "Always do it yourself" ignores that FDA review is genuinely unforgiving of avoidable mistakes, and the cost of a mistake is measured in months, not dollars.
Two variables do most of the work in the actual decision:
Submission frequency. One device, one filing, is a different problem from a five-year pipeline of Class II devices sharing a platform. Frequency is what makes in-house investment pay off — the second and third submissions get faster once the first one builds real institutional knowledge.
Predicate clarity. If your device sits squarely in a well-trodden product code with dozens of cleared predicates and no novel technological characteristics, the substantial equivalence argument mostly writes itself. If your predicate choice is contestable — a borderline intended use, a genuinely new feature, a product code with sparse or dated predicates — this is precisely where judgment earns its keep, because FDA's questions will land exactly on the ambiguity you glossed over. This is also where a CE-marked device carries hidden complexity: what actually transfers from a European file to a 510(k) is rarely obvious from the technical file alone.
Timeline pressure and in-house RA depth modify both, but frequency and predicate clarity are the two questions worth answering honestly before anything else.
What genuinely belongs in-house, regardless of your answer
Some things stay in-house no matter which way you lean, and it's worth saying so plainly rather than implying everything should be handed off.
Your device description and intended use. Nobody outside your company understands your device as well as your engineering and clinical team. A consultant translates that understanding into FDA's language; they cannot originate it.
Ownership of your quality management system. Under the QMSR, effective since February 2026, your QMS is your responsibility, not a vendor's. Documentation support can be outsourced. Ownership cannot.
Institutional memory across submissions. If you're a repeat filer, every submission should make the next one faster. That only happens if someone on your team is actually building that knowledge, not routing it entirely through an outside firm each time.
Where inexperience actually costs time, not just money
This is the part first-time filers underestimate, and it's mechanical rather than a matter of opinion.
Since October 1, 2023, 510(k)s must be submitted electronically through eSTAR. eSTAR changed the front-end screening process in a way a lot of competitor content still glosses over: a properly completed eSTAR submission is not expected to go through the old Refuse to Accept process at all. Instead, FDA runs a technical screening, checking that the file is complete and uncorrupted, generally within 15 calendar days. If it fails, the mechanics are similar to a classic RTA hold — you get 180 days to submit a complete replacement before the file is considered withdrawn — but the criteria are about completeness, not the older administrative checklist. If you want the detailed failure-mode breakdown, we've covered it separately: what actually causes an RTA or technical screening hold, and the ten most common reasons FDA submissions fail.
Passing that first screen is table stakes. The real risk sits in substantive review, and here the mechanics matter more than most first-time filers realize. Under MDUFA V, FDA's own performance goals commit to a Substantive Interaction — the point where FDA communicates its first round of questions — within 60 FDA days for 95% of submissions, and a full MDUFA decision within 90 FDA days for 95% of submissions.
"FDA days" is not the same as calendar days, and this is where timelines quietly blow up. The clock pauses whenever FDA is waiting on you — most commonly during an Additional Information request. A 90-FDA-day goal can stretch to six months or more of real elapsed time if your response to an AI request takes weeks to prepare, because the delay is on your side of the clock, not FDA's. Average total time to decision, measured in calendar days across all submissions, has run in the range of roughly 110–130 days in recent fiscal years — and that average already includes the submissions that sailed through cleanly. A submission with a contestable predicate or a poorly framed substantial equivalence argument doesn't miss the goal by a little. It generates one or more AI cycles, each adding weeks, on top of the time needed to actually solve the underlying problem.
This is the asymmetry that should drive the decision more than any fee comparison: a bad predicate call caught by FDA doesn't cost you a consultant's invoice. It costs a review cycle, and depending on how fundamental the problem is, sometimes a second submission. Choosing the right predicate device is the single highest-leverage judgment call in the entire process, and it's also the one most dependent on having seen how FDA has actually responded to similar arguments before — which is exactly the kind of pattern recognition that's hard to build on a single submission.
A decision framework
| Factor | Leans in-house | Leans toward outside expertise |
|---|---|---|
| Submission frequency | Repeat filer, ongoing pipeline | One-off or first-ever filing |
| Predicate clarity | Well-established product code, clear predicates | Contestable, novel, or sparse predicate landscape |
| In-house RA depth | Someone has done this before | No regulatory affairs background on staff |
| Timeline pressure | Flexible, no hard external deadline | Investor, distributor, or launch deadline attached |
| Consequence of delay | Manageable | High — a missed window has real commercial cost |
| Clinical data involved | None, or standard bench/bio testing only | Any clinical investigation, human factors, or novel technology |
Most real companies don't land cleanly on one side. That's the honest answer, not a hedge.
The hybrid most experienced manufacturers land on
The false binary in "in-house vs. outsourced" assumes the whole submission moves as one unit. It doesn't have to, and in practice the strongest setups split it: predicate strategy and the substantial equivalence argument reviewed by someone with direct FDA submission experience, while the underlying documentation, testing coordination, and submission assembly stay owned by your team.
That's a genuinely different shape from either full outsourcing or going it alone, and it's why Complizen runs two lines rather than one. For teams that want to keep the work in-house but need structure and traceability rather than a blank Word document, the Superagent platform organizes the submission against FDA's own databases and guidance so nothing is built from a standing start. For teams that want the predicate judgment and risk transferred to someone who has made this exact call before, the 510(k) submission service is reviewed throughout by regulatory experts with direct FDA submission experience. Which one fits depends on where you land in the framework above — not on which is being sold harder.
What it actually costs — structurally, not in numbers
Specific fee ranges vary too much by device complexity and firm to be useful here, so this is about structure, not a number to anchor on.
In-house has a real cost even when it looks free. Hiring a competent regulatory affairs professional takes months, and their first submission at your company is rarely their fastest — ramp time is a genuine cost, not a rounding error. If you're a one-time filer, that ramp cost is being spent on a single use.
Outside expertise is usually priced by scope, not by the hour of a single task. Project-based and retainer structures are both common; hourly work is less common for full submissions specifically because the value is in judgment calls made early, not hours logged late.
The number that actually matters is the cost of getting it wrong, not the cost of getting help. An AI cycle costs weeks. A rejected predicate argument that forces a resubmission costs months and, depending on your fee timing, can trigger FDA's small business re-qualification and fee rules a second time. Weigh outside help against that number, not against the invoice in isolation.
Common mistakes
Assuming a first hire replaces years of pattern recognition. A single junior regulatory affairs hire, however capable, hasn't seen how FDA responds to your specific product code's edge cases yet. That's not a knock on the hire — it's just not something a first year builds.
Outsourcing everything, including QMS ownership. Documentation support can come from outside. Owning your quality system cannot be fully delegated, and treating it as someone else's problem creates gaps that surface at inspection, not at submission.
Choosing based on price alone. The cheapest option that gets your predicate wrong is not actually cheap once you count the review cycle it costs you.
Waiting until after a technical screening hold to bring in expertise. By then you've already spent the 15-day screening window and possibly some of your 180-day cure period. Judgment is worth the most before submission, not after a hold notice.
Assuming the 90-day goal is a guarantee. It's a performance goal for 95% of submissions, measured in FDA days that pause during your own response time. Building a launch plan around exactly 90 calendar days is a common and avoidable planning error.
Frequently asked questions
Do I need a consultant for my FDA 510(k) submission? Not always. If you're a repeat filer with a clear predicate, in-house regulatory affairs depth, and no hard deadline, running it in-house is a reasonable choice many manufacturers make well. The case for outside expertise strengthens sharply for a first submission, a contestable predicate, or a tight timeline — because a wrong judgment call costs a review cycle, not just a fee.
Can I submit a 510(k) myself without a consultant? Yes. There's no legal requirement to use a consultant or law firm for a 510(k) submission. The practical question isn't whether you're allowed to — it's whether your predicate selection and substantial equivalence argument are strong enough to survive FDA's substantive review without avoidable AI cycles.
What does a 510(k) consultant actually do that I can't do myself? The highest-value contribution is usually predicate selection and framing the substantial equivalence argument — judgment calls informed by having seen how FDA has responded to similar arguments before. Documentation assembly, testing coordination, and eSTAR completion are more mechanical and more reasonably kept in-house, particularly with the right software support.
Is it cheaper to do my 510(k) in-house? Often, if you're a repeat filer with a clear predicate. The comparison changes for a one-time filer, because the ramp-up cost of a first hire or a first attempt is a real cost even without an invoice, and a rejected predicate argument costs far more in time than any advisor's fee.
What happens if I choose the wrong predicate device myself? FDA's substantive review will surface it, typically through an Additional Information request or, in more serious cases, a Not Substantially Equivalent determination. Either extends your timeline well beyond the 90-FDA-day goal, since the review clock pauses while you prepare a response. It does not automatically mean a full restart, but it does mean lost time you were likely not planning for.
Do I need special software to use eSTAR, or do I need a consultant? Neither is strictly required — eSTAR is a free FDA-provided template. What determines the outcome is whether the content going into it is well-organized and complete, not the tool itself. Software built around FDA's own databases and guidance can meaningfully reduce the documentation burden without requiring you to hand off the judgment calls.
How many 510(k)s do I need to file before in-house expertise makes sense? There's no fixed number, but the logic is about amortization: institutional knowledge from a first submission only pays off if there's a second and third to apply it to. A single-device company evaluating a one-time filing is in a different position from a manufacturer building a five-year pipeline on a shared platform.
What's a hybrid approach between in-house and fully outsourced? Keeping documentation, testing coordination, and submission assembly in-house — often supported by software built around FDA's own data — while having predicate selection and the substantial equivalence argument reviewed by someone with direct FDA submission experience before filing. This splits the parts of the work that benefit from institutional ownership from the part that benefits most from pattern recognition.
Does FDA's 90-day review goal mean my submission takes 90 days total? No, and this is a common and costly misreading. The 90-FDA-day goal (achieved for 95% of submissions under MDUFA V) measures days FDA is actively reviewing your file. The clock pauses during any period FDA is waiting on you, such as an Additional Information response. Total calendar time to decision has averaged meaningfully higher than 90 days across recent fiscal years, and any submission that generates an AI cycle will run longer still.
Should first-time filers always use a consultant? Not always, but the case is strongest here. A first-time filer has no institutional pattern recognition to draw on, which is precisely the gap outside expertise closes fastest. If your predicate is genuinely unambiguous and your timeline has slack, careful in-house work can still succeed — but the margin for error is thinner than it looks.
Key takeaways
The decision turns on frequency and predicate clarity, not budget alone. A repeat filer with a clear predicate and in-house depth can genuinely succeed alone. A first-time filer with an ambiguous predicate is taking on real, quantifiable risk.
eSTAR changed the front door. Since October 2023, compliant eSTAR submissions face technical screening rather than the classic RTA process — faster and more mechanical, but unforgiving of incomplete sections.
"FDA days" and calendar days are not the same thing. The 90-FDA-day MDUFA V goal pauses during any period FDA is waiting on you. Planning around 90 calendar days is a common, avoidable mistake.
The real cost of a wrong call is a review cycle, not a fee. That asymmetry — weeks or months of delay against the cost of getting expert input up front — should drive the decision more than any price comparison.
Most manufacturers land on a hybrid. Predicate judgment from someone who has done this before, documentation and ownership kept in-house. That split, not a binary choice, is what the strongest setups actually look like.
Complizen helps international medical device manufacturers reach FDA 510(k) clearance, combining a software platform for in-house regulatory teams with full-service consultancy for teams without in-house FDA expertise.
If your predicate situation is anything less than clear-cut, that's the single judgment call worth getting a second opinion on before you file. Complizen's 510(k) submission service is built around exactly that: predicate strategy and substantial equivalence framing reviewed by regulatory experts with direct FDA submission experience. See how the submission service works →
References
- FDA — 510(k) Submission Process. https://www.fda.gov/medical-devices/premarket-notification-510k/510k-submission-process
- FDA — Refuse to Accept Policy for 510(k)s (guidance, reissued April 21, 2022). https://www.fda.gov/regulatory-information/search-fda-guidance-documents/refuse-accept-policy-510ks
- FDA — MDUFA Performance Goals and Procedures, Fiscal Years 2023 Through 2027 (MDUFA V Commitment Letter). https://www.fda.gov/media/73507/download
- FDA — eSTAR Program. https://www.fda.gov/medical-devices/how-study-and-market-your-device/estar-program
- FDA — MDUFA Performance Reports. https://www.fda.gov/about-fda/user-fee-performance-reports/mdufa-performance-reports
- Federal Register — Medical Device User Fee Rates for Fiscal Year 2027. https://www.federalregister.gov/documents/2026/07/30/2026-15335/medical-device-user-fee-rates-for-fiscal-year-2027
- FDA — Quality Management System Regulation (QMSR). https://www.fda.gov/medical-devices/postmarket-requirements-devices/quality-management-system-regulation-qmsr
- FDA — Acceptance Checklists for 510(k)s. https://www.fda.gov/medical-devices/premarket-notification-510k/acceptance-checklists-510ks