What Do FY2026 USPTO Patent Grant and Publication Trends Show?

The USPTO issued 315,825 utility patents in FY2026, down 3.8% from FY2025, according to Patently-O's FY2026 data update. Roughly half of that decline is a calendar artifact: FY2025 contained 53 Tuesdays (the day utility patents issue), while FY2026 returned to the usual 52. Even after setting that aside, the weekly grant rate was down about 2%. The more substantive shifts are in the mix of what issued — fewer continuing applications, and a first-time drop below 40% for patents whose first-named inventor claims US residence.

The headline numbers

The 3.8% figure is the raw year-over-year change. Because one fewer issue day mechanically removes a week's worth of grants, the underlying weekly rate — about 2% lower — is the better signal of whether examination output actually slowed.

Why the continuing-application drop matters

Continuing applications here means continuations, divisionals, and continuations-in-part (CIPs) — filings that spin off from an earlier pending or issued application rather than starting fresh.

A fall from 36% to roughly 26% of all grants is a large single-year move. It means a smaller share of issued patents trace back to an earlier family member, and a larger share are original filings. That shift is worth watching because continuations are a common tool for:

• Keeping a family alive while narrowing or broadening claims
• Pursuing different claim scope after an initial rejection or allowance
• Preserving the option to file new claims as a technology or market develops

If the share keeps falling, it suggests either that applicants are choosing to let families lapse rather than extend them, or that examination timing and cost are pushing filers toward single-shot prosecution. The data update does not establish which, so treat the cause as open.

US-resident inventors falling below 40%

For the first time, patents whose first-named inventor claims US residence made up less than 40% of grants. Two cautions before reading too much into it:

1. First-named inventor is a proxy, not a count of US inventorship. A patent with a US first-named inventor and foreign co-inventors counts as US here; a patent with a foreign first-named inventor and US co-inventors does not. The metric measures where the lead inventor sits, not the full team.
2. The denominator moved too. Total grants fell, so a stable foreign-origin count would still raise the foreign share. A falling US share does not by itself prove US-origin filings declined in absolute terms.

What the new data cuts add

The update adds two views beyond raw grant counts:

• Published applications by origin — shows where filings come from before they issue, which leads grants by roughly the publication lag.
• Cross-border inventor teams — captures patents with inventors in more than one country, a category that a single-origin count misses entirely.

Together these matter because grant counts alone conflate at least three things: how much is being filed, how much is being allowed, and how long examination takes. Publication data reflects filing activity earlier in the pipeline; cross-border team data shows whether innovation is increasingly collaborative across borders even as single-country shares shift.

How to use these trends

• If you are tracking filing activity: weight published-application data more heavily than grants, since grants lag filings by years and are distorted by issue-day counts.
• If you are planning a continuation strategy: note that continuations are a shrinking share of grants industry-wide, but the data does not say whether that reflects strategy, cost, or examiner behavior in your technology area.
• If you are assessing US vs. foreign innovation: use cross-border team data alongside the first-named-inventor figure, because the 40% threshold is sensitive to how inventorship is attributed.
• If you need the full post: the FY2026 update is a paid Patently-O post; the excerpt above is the free portion, and full access requires membership or login.