Guide
FDA 510(k) Clearance: A Practical Readiness Guide
An FDA 510(k) Premarket Notification under section 510(k) of the Food, Drug, and Cosmetic Act and 21 CFR Part 807 Subpart E is the dominant route to the US medical device market by volume — roughly 3,000 clearances per year versus around 50 PMA approvals. A 510(k) demonstrates that the new device is substantially equivalent to a legally marketed predicate device, matching intended use and either matching technological characteristics or showing that different characteristics do not raise different questions of safety and effectiveness. The MDUFA V performance commitment is a 90 FDA-day review clock; the real-world median is around 177 calendar days once Additional Information hold cycles are counted. This guide is for regulatory affairs leads, R&D managers, and quality leads at medical device manufacturers preparing a 510(k) submission. It walks through the predicate-selection decision that drives everything downstream, the eSTAR mandatory submission format, the Refuse-to-Accept (RTA) checklist that decides whether your submission even enters substantive review, the performance-data expectations, and a readiness path that materially shortens time to clearance.
Start free trial Free trial, no credit card, onboard in days, not months.
Predicate selection — the decision that drives everything
The predicate is the legally marketed device against which substantial equivalence is established. It must have the same intended use as the new device, and either the same technological characteristics or different characteristics that do not raise different questions of safety and effectiveness. Predicate selection drives the entire submission: the performance data the new device must produce, the labelling claims that are defensible, the comparison table that anchors the SE argument, and the review division that handles the submission. Strong submissions pick a single primary predicate with optional reference devices for specific characteristics; weak submissions split-reference across multiple predicates and produce a hybrid intended use that FDA rejects as 'no clear predicate'. Predicates withdrawn for safety reasons cannot be used as predicates regardless of their cleared status.
eSTAR — the mandatory submission format
Since 1 October 2023, FDA requires 510(k) submissions in the eSTAR electronic format. eSTAR is a guided PDF template that enforces structure, embeds the RTA checklist, dynamically opens and closes sections based on device type, and produces a submission package FDA's Document Control Center can ingest without manual indexing. Paper and legacy eCopy submissions are no longer accepted for 510(k) (the only exceptions are dual submissions and very narrow categories). eSTAR has materially reduced RTA hold rates because the template will not let a sponsor close a section with missing content. Manufacturers preparing their first eSTAR submission should download the current template version from FDA early and rehearse the workflow on a partial draft, because the in-template logic is non-obvious and changes between versions.
Refuse to Accept (RTA) — the first gate
Within 15 calendar days of receipt, FDA runs the RTA checklist against the submission. If any RTA item is missing or inadequate, the submission is refused, the user fee is forfeited or partially refunded depending on circumstance, and the sponsor must address the gaps and resubmit. The RTA checklist covers administrative completeness (user fee, eCopy or eSTAR format, basic identifying information) plus substantive completeness (device description, indications for use, predicate identification, substantial equivalence discussion, performance data, labelling, sterilisation, biocompatibility, software documentation if applicable). The single most common RTA failure is performance-data sections that reference tests without including the actual reports. eSTAR has reduced RTA failures but not eliminated them — the template enforces presence, not adequacy.
Performance data — what FDA actually expects
The performance-data section is where SE is won or lost. The expectations scale with risk: bench testing per recognised consensus standards (FDA maintains a Recognized Consensus Standards Database — using a recognised standard with a Declaration of Conformity shortens review); biocompatibility per ISO 10993 series; sterilisation validation per the appropriate ISO standard (11135, 11137, 17665, etc.); software documentation per FDA's 2023 software guidance with the documentation level driven by the device's Basic Documentation Level vs Enhanced Documentation Level classification; cybersecurity per FDA's 2023 cybersecurity guidance including SBOM, threat model, and security risk assessment; usability per IEC 62366; clinical data when needed (rare for traditional 510(k), more common for AI-enabled and novel-technology devices).
MDUFA V timelines and AI-NAI hold cycles
MDUFA V (FY2023-2027) sets the 90 FDA-day SE/NSE decision goal. The clock only runs while the submission is with FDA — when FDA issues an Additional Information (AI) request, the clock stops and resumes when the sponsor responds. AI requests are routine: roughly 70-80% of 510(k)s receive at least one AI hold during review. The number of AI cycles is the single biggest determinant of real-world calendar time to clearance. Each AI cycle adds 30-180 calendar days. Strong submissions reduce AI cycles by anticipating the questions: the most common AI categories are performance test gaps, software documentation depth, cybersecurity SBOM completeness, biocompatibility justifications, and labelling claims that exceed the supporting data. Median time to clearance under MDUFA V is around 177 calendar days; the top decile manufacturers achieve 110-130 days.
Special, Abbreviated, and Traditional 510(k)
Three submission types. Traditional — the default, used when none of the other types fit; full SE discussion and performance data. Abbreviated — used when the device conforms to recognised consensus standards or FDA guidance documents; the submission relies on declarations of conformity instead of full test data summaries; common for devices in well-characterised categories. Special — used only when the sponsor is modifying their own already-cleared device, the modification does not affect intended use or fundamental scientific technology, and design controls have been followed; FDA review goal is 30 FDA-days instead of 90. Picking the wrong type wastes time — Special 510(k)s submitted for changes that affect fundamental technology are converted to Traditional and lose 60+ days.
A 120-day 510(k) readiness path
Days 1 to 15: predicate selection locked; intended use statement finalised; SE strategy documented; FDA Q-Submission (pre-sub) meeting requested if novel technology. Days 16 to 50: design history file consolidated; performance test plan executed against recognised standards; biocompatibility, sterilisation, software documentation, and cybersecurity artefacts assembled. Days 51 to 85: eSTAR template populated; comparison table to predicate completed; labelling drafted including indications for use and any caveats; usability and human factors evidence packaged. Days 86 to 110: internal regulatory review with RTA pre-check; pre-submission meeting feedback incorporated if applicable; final QA review and signature workflow. Days 111 to 120: user fee paid; eSTAR submitted; CDRH receipt confirmation tracked; AI response capacity reserved for the expected 30-90 day AI cycle.
Where this lives in V5 Ultimate
The clauses above aren't theoretical — every one maps to a shipped module and an industry profile. Jump to the parts of the product that turn this guide into evidence on a Monday morning.
Modules that do the work
Quality management
21 CFR 820 / QMSR design controls feeding the 510(k) submission.
Document control
Design history file, performance test reports, and submission artefacts under change control.
Audits & CAPA
Post-market surveillance feeding the next submission's performance evidence.
Compliance self-assessment
RTA pre-check and submission-readiness scoring.
Industries this hits hardest
Frequently asked
Is 510(k) clearance the same as FDA approval?
No. 510(k) clearance is a substantial-equivalence determination — FDA agrees the device is at least as safe and effective as the predicate. FDA approval is the term reserved for PMA, De Novo, and HDE pathways where FDA makes an independent safety-and-effectiveness determination. Labelling a 510(k)-cleared device as 'FDA-approved' is a misbranding violation under the FD&C Act.
How much does a 510(k) cost?
User fees only: the FY2026 standard 510(k) fee is around $24,335, with small business reduction available (around $6,084 for qualifying applicants). Real cost is mostly the performance-data programme — bench testing, biocompatibility, sterilisation, software V&V, cybersecurity, usability — which typically runs $150K-$500K for a Class II device depending on complexity. Regulatory consulting fees add $50K-$200K for sponsors without in-house regulatory.
Can we use a predicate that was withdrawn from the market?
It depends on why it was withdrawn. A predicate withdrawn for safety reasons cannot be used as a predicate (21 CFR 807.100(b)). A predicate withdrawn for commercial reasons (manufacturer discontinued, market consolidation) generally remains usable. FDA's 510(k) database flags safety-related withdrawals. When in doubt, raise the question at a Pre-Submission meeting before committing the submission strategy.
Do we need clinical data for a 510(k)?
Most 510(k)s clear on bench performance and biocompatibility without clinical data. Clinical data is typically required when: bench testing cannot adequately characterise safety or effectiveness; the device has novel technological characteristics; FDA guidance for the specific device category calls for clinical data; or the device uses AI/ML and clinical performance evidence is the only way to demonstrate intended use. The Pre-Submission process is the right place to confirm clinical data expectations before committing.
See it on your shop floor.
Free trial, no credit card, onboard in days, not months.
Related reading
Editorial·Reviewed against the V5 marketing knowledge base. Spot something off? .