De Novo Classification Request for Novel Medical Devices

1. Key takeaways

A De Novo request applies to novel devices whose risks are controllable but for which no lawful predicate exists. Under FD&C Act §513(f)(2), the request asks FDA to classify the device directly into Class I or Class II (with special controls where appropriate), so that the PMA route can be avoided. FDA has set a review goal of 150 days for De Novo requests.

• The De Novo evidentiary focus differs fundamentally from the 510(k): with no predicate, the sponsor must show directly that general controls (Class I) or general controls plus special controls (Class II) adequately manage the risks.
• A device granted De Novo authorization becomes a lawful predicate for future 510(k)s, so an early De Novo success creates a lasting competitive barrier.
• Use of eSTAR is mandatory for De Novo submissions, except where an exemption has been granted.
• The FY2026 De Novo fee of USD 173,782 (small business USD 43,446) sits well above the 510(k) fee but below the PMA fee.
• Risk management documentation carries far greater weight in a De Novo dossier than in a 510(k).
• The predicate search record should be archived, since it justifies why the De Novo route, rather than a 510(k), was selected.

2. Legal basis

The table below lists the authorities governing the De Novo route, including the classification mechanism, the FY2026 fee schedule, and the eSTAR mandate, each verified against official sources in 2026.

3. Expert insights

Insight 1: The De Novo evidentiary center of gravity differs from the 510(k). A 510(k) demonstrates substantial equivalence to a predicate; a De Novo has no predicate, so the sponsor must demonstrate directly that risk is adequately controlled by general controls (Class I) or by general controls plus special controls (Class II). In practical terms, this means that the risk management documentation, typically built on ISO 14971, carries far greater weight in a De Novo dossier than in a 510(k). Each identified risk must be mapped to a control, and each control must be mapped to evidence that the control works. Dossiers organized this way answer the reviewer’s central question before it is asked.

Insight 2: A successful De Novo creates a new predicate. A device granted De Novo authorization may, under FD&C Act §513(f)(2), subsequently serve as a predicate for other 510(k) submissions. For a first-time market entrant, the De Novo is therefore both a cost and a moat: the company that walks the path first establishes the reference point on which all later competitors will depend. This strategic dimension should inform the decision of whether to invest in a De Novo now or wait.

Insight 3: eSTAR is mandatory for De Novo. According to the FDA eSTAR Program page, all De Novo submissions must use the eSTAR template unless an exemption is obtained. When eSTAR versions iterate, retirement dates are set for older versions (for example, version 6.2 was retired on 2026-08-03), so the current version must be checked before submission. Building the dossier in a version close to retirement risks a rejection on formal grounds at the moment of filing.

Insight 4: The De Novo fee is significantly higher than the 510(k) fee but lower than the PMA fee. The FY2026 standard fees are: 510(k) USD 26,067, De Novo USD 173,782, and PMA USD 579,272. The gaps between these figures reflect the differences in review resources consumed, and sponsors should weigh fee and timeline together when selecting the pathway rather than treating cost as the sole determinant.

Because the De Novo sits between the 510(k) and the PMA in cost, review intensity, and evidentiary expectation, it rewards early planning more than either of the other routes. The pathway decision should be taken while the development program is still flexible enough to generate the evidence the classification argument will need. Waiting until the data package is complete to discover that the classification story is weak leaves little room to add the risk-control evidence that a Class I or Class II argument requires, and the resulting review cycles consume both budget and market-entry time.

Teams preparing their first De Novo should also align internal reviewers with this pathway logic. Colleagues accustomed to 510(k) work tend to draft content as a comparison narrative, which has no anchor here. Briefing the writing team on the risk-control-evidence structure before drafting begins prevents weeks of restructuring later and produces a dossier whose sections map cleanly onto the questions the reviewing division will ask.

4. Practical recommendations

• Complete the predicate search first and retain the evidence; only move to De Novo after the absence of a lawful predicate is confirmed on the record.
• Organize the dossier around risk management as the main thread, with risk, control, and evidence mapped in three aligned columns.
• Check the current eSTAR version and the retirement dates of older versions before submission.
• Use the 150-day De Novo review goal as the scheduling baseline and reserve buffer for Additional Information requests.
• Build the budget on the FY2026 fee of USD 173,782 (small business USD 43,446) and leave room for fiscal-year increases.

5. Verification checklist

• □ Predicate search completed and the absence of an available predicate confirmed and evidenced.
• □ Intended classification (Class I or Class II) determined.
• □ Risk analysis document completed, with risk, control, and evidence columns closed.
• □ If Class II is requested, proposed special controls content prepared.
• □ Dossier prepared using the current version of eSTAR.
• □ De Novo fee budgeted at the FY2026 rate.

How GRZAN can help. GRZAN provides medical device and IVD registration, quality system, US Agent and local testing support for the United States market. Our team manages FDA establishment registration and listing, 510(k)/De Novo/PMA submissions, and supports post-market reporting obligations. Business enquiries: info@grzanglobal.com