The baseline: device decisions by pathway
[Chart: openFDA via PharmaDossier (2019-2025, snapshot 6 Jul 2026); BTIG via MD+DI (2026 projection). See the live document for the chart.]
From 2019 to 2024, original PMAs for high-risk devices ran at 22 to 35 a year. They rose to 40 in 2025, and BTIG projects about 51 for 2026. Other pathways were steadier:
| Pathway | 2025 | 2026 (projection) |
|---|---|---|
| 510(k) clearances | about 3,200 to 3,240 | about 3,239 |
| De Novo grants | 27 | 27 |
| Novel devices (FDA count) | 124 | not yet reported |
| All marketing authorizations (FDA) | 5,640 (vs 5,564 in 2024) | not yet reported |
The 2025 figures are from FDA’s CDRH annual report via Emergo by UL. The 2026 projections are BTIG’s.
Two points about the baseline:
- Counts differ by source. openFDA shows 40 PMAs for 2025; FDA’s annual report says 41.
- Volume is up, speed is down. In the first half of 2026, the average time to PMA approval was about 599 days, and 510(k) reviews took about 156 days.
Leading indicators: startups, Breakthrough designations, and AI
Fewer, larger startup rounds. J.P. Morgan’s medtech venture data:
| Period | Venture funding | Rounds | Seed and Series A rounds |
|---|---|---|---|
| 2024 | about $12.6B | 415 | 170 |
| 2025 | $11.3B | 323 | 113 |
| 2026, Jan to Sep | $8.0B | 278 | not reported |
Round counts fell by about 22% in 2025. Seed and Series A funding of $1.9 billion matched 2020 as the lowest level in J.P. Morgan’s six-year series. PitchBook calls this the “medtech VC paradox”: strong headline dollars, but a thinner pool of early companies. That pool is what feeds future PMA and De Novo filings. Through Q3 2026, rounds are running ahead of 2025’s pace (278 in nine months against 323 for all of 2025), which suggests some recovery.
A steady flow of Breakthrough devices. These are the novel devices most likely to need a PMA or De Novo.
| Fiscal year | Breakthrough designations |
|---|---|
| FY2021 | 206 (peak) |
| FY2022 | 166 |
| FY2023 | 145 |
| FY2024 | 165 |
| FY2025 | 164 |
As of 30 June 2026, FDA had granted 1,320 designations in total and authorized 210 of those devices, about 16%. The gap means a large stock of designated devices could still reach the market.
AI is the fastest-growing segment. FDA’s AI-enabled device list grew from 235 new authorizations in 2024 to 335 in 2025, and reached 1,614 in total by June 2026. About 75% are radiology. PitchBook notes that close to three-fifths of capital in the top medtech deals went to AI-enabled companies, so startup money is flowing toward the segment that is already growing fastest.
Forecast: 2026 to 2030
How we built it. We started from each pathway’s recent level, then adjusted for the indicators above. These are judgment calls, not a statistical model.
AI-enabled devices: the clearest growth
[Chart: Actuals: FDA AI-Enabled Medical Device List via IntuitionLabs (list of 4 Sep 2026). Forecast: AskDona editorial estimate, October 2026. See the live document for the chart.]
AI authorizations grew about 22% a year from 2023 to 2025. Our base case slows that to about 12% a year, reaching about 560 in 2030.
High-risk and novel devices
| Measure | 2025 actual | 2026 estimate | 2027 to 2030 a year: low / base / high |
|---|---|---|---|
| Original PMA approvals | 41 (FDA) | about 51 (BTIG) | 35 to 40 / 45 to 50 / 55 to 60 |
| De Novo grants | 27 | about 27 (BTIG) | 22 to 25 / 28 to 32 / 35 to 40 |
| 510(k) clearances | about 3,200 | about 3,239 (BTIG) | about 3,100 / 3,250 to 3,400 / about 3,600 |
The reasoning behind the base case:
- The Breakthrough stock supports PMA and De Novo volume. More than 1,100 designated devices are not yet authorized.
- Weaker early-stage funding should trim new filings late in the decade. Seed and Series A rounds fell by a third in 2025, and in our judgment many of those companies would have filed toward the end of the decade. So we expect PMAs to settle below the 2026 peak rather than keep rising.
- AI growth continues. Venture money concentrates in AI-enabled companies. The CDRH guidance list for FY2027 includes a final guidance on AI-enabled device software functions, which should make submissions more predictable.
In the low case, longer review times and the MDUFA VI transition slow decisions. In the high case, the 2026 funding recovery holds and AI guidance speeds up clearances.
Risks to the forecast
- Review capacity. FDA cut more than 3,000 roles in 2025. In the first half of 2026, the average time to PMA approval rose to about 599 days.
- MDUFA VI. MDUFA V expires on 30 September 2027, and new goals and fees apply after that.
- Lost Breakthrough payment incentive. Devices designated after 30 September 2026 lose the alternative Medicare payment route, which may reduce designations.
- Counting differences. openFDA, FDA annual reports, and trade trackers report different totals for the same year.
Using this forecast in AI and RAG
- Index FDA databases and the CDRH annual report as the ground truth for counts.
- Store each indicator with its source and date. Funding trackers differ.
- Label forecast figures as estimates, with their scenario.
- Refresh twice a year: when BTIG and similar mid-year counts appear, and when the CDRH annual report is published.
Common mistakes
- Reading strong headline venture dollars as more startups. Round counts fell.
- Treating every Breakthrough designation as a future approval.
- Comparing counts from different databases.
- Extrapolating the 2026 PMA jump straight forward.
Checklist
- Counts tied to FDA sources
- Pathways kept separate (PMA, De Novo, 510(k))
- Funding and designation data dated
- Scenario labels on all estimates
- Refresh schedule set
How AskDona helps
AskDona checks your market models, investor decks, and regulatory plans against the FDA data and industry reports you load. It shows the source behind every number and flags any assumption without one.
Run a sample device forecast evidence check
AskDona shows the evidence; your strategy and regulatory teams decide.
Editor’s note. This article is general information, not investment advice. Historical counts come from FDA data and the sources named in the references. Funding data comes from J.P. Morgan and PitchBook. The forecast is our own estimate as of 30 October 2026.
FAQ
How many devices does FDA approve through PMA each year? Between 22 and 35 from 2019 to 2024, and 40 to 41 in 2025. BTIG projects about 51 for 2026.
Is the number of medtech startups growing? Venture dollars are high, but the number of rounds fell about 22% in 2025, and seed and Series A rounds fell by a third. 2026 shows some recovery.
How many AI-enabled devices will FDA authorize? Our base case is about 410 in 2027, rising to about 560 in 2030.
Will startup growth raise PMA approvals? Modestly. The stock of Breakthrough devices supports volume, but weaker early-stage funding should trim filings late in the decade.
References
All pages opened and checked by October 30, 2026.
| Source | What it supports |
|---|---|
| PharmaDossier: FDA device numbers 2026 (openFDA) | PMA and 510(k) counts, 2019 to 2026 H1 |
| Emergo by UL: CDRH 2025 annual report | 2025 PMA, De Novo, novel devices, totals |
| MD+DI: FDA approval timelines in 2026 | BTIG projections and review times |
| MedTech Dive: More devices, taking longer | Review times, FDA staffing |
| J.P. Morgan: Q4 2025 medtech report (PDF) | 2024 and 2025 venture, seed and Series A |
| J.P. Morgan: Q3 2026 deal reports | $8.0B and 278 rounds through Q3 2026 |
| PitchBook: The Medtech VC Paradox (PDF) | Early-stage decline, AI share |
| FDA: Breakthrough Devices Program | 1,320 designations, 210 authorizations |
| MedTech Dive: Breakthrough program FY2026 | Designations by fiscal year |
| IntuitionLabs: FDA AI-enabled device list | AI authorizations by year |
| Avania: MedTech Industry Pulse, July 2026 | MDUFA VI timing, Breakthrough payment change |
| FDA Law Blog: CDRH FY2027 guidances | AI guidance priorities |