Measured in the I-129 petition microdata, not announced schedules.
USCIS runs an additional H-1B selection round when the petitions filed from the first round will not fill the 85,000 cap. It has happened in 3 of the 4 cap seasons covered by our petition data, and it leaves an unmistakable trace: a second wave of petition receipts months after the initial filing window closed.
USCIS selects more registrations than the 85,000 cap because not every selected registrant files a petition. When the petitions that do arrive will not fill the cap, it selects again from the registrations already submitted that March. There is nothing to re-submit and no second registration period: a registration that was not selected the first time stays in the pool for the rest of that cap season, and anyone newly selected is notified in their USCIS online account.
| Cap FY | Initial window April–July |
After it closed August onward |
Share arriving later | Later wave? |
|---|---|---|---|---|
| FY2021 | 86,979 | 12,601 | 12.7% | Yes |
| FY2022 | 65,008 | 24,513 | 27.4% | Yes |
| FY2023 | 91,777 | 37 | 0.0% | No |
| FY2024 | 58,217 | 33,646 | 36.6% | Yes |
Counts are cap-subject petitions that were selected and filed, bucketed by the date USCIS received them. The split is sharp rather than borderline: July is the trough in every season, and the seasons that ran an additional round show tens of thousands of receipts in August and later while the seasons that did not show a few hundred. This measures arrival dates, which is the trace a later round leaves, rather than an announcement of one.
It does not predict one. Whether USCIS runs another round for FY2027 turns on how many selectees filed during the initial window — a count only USCIS holds, and one it acts on before publishing. What the table above gives is the base rate: how often an additional round has happened in the seasons we can measure, and how much of a season came through it when it did. Treat anyone quoting a probability for the current season as quoting a guess.
Our petition data covers FY2021–FY2024 as a frozen FOIA snapshot, so seasons after it are outside what this page can measure.
There is in some cap seasons. USCIS selects again from the registrations already submitted when the petitions filed from the first round will not fill the 85,000 cap. In FY2024, 36.6% of that season's cap petitions were received after the initial filing window closed — 33,646 petitions. Seasons with no additional round look completely different: FY2023 shows 0.0%. It is decided after USCIS sees how many selectees actually filed, so it is not announced in advance.
Nothing published lets anyone answer that before USCIS does: the decision depends on how many FY2027 selectees filed during the initial window, which USCIS alone can count. What the history below shows is how often it has happened and how large the later waves were when it did. If a round is run, registrants who were not selected the first time are notified in their USCIS online account; there is nothing to re-submit.
No. Additional rounds draw from the registrations already submitted that March. A registration that was not selected stays in the pool for the rest of that cap season.
Each bar is the share of a cap season's selected-and-filed I-129 petitions that USCIS received after the initial filing window closed, from the FOIA-released petition microdata. It measures when petitions arrived, which is the trace an additional round leaves, rather than an announcement of one.
Cap season calendar · Selection odds by year
Registration and selection totals are USCIS's own published H-1B electronic registration statistics; the change of selection basis is the FY2025 beneficiary-centric rule. Petition receipt dates come from the I-129 microdata released under the Freedom of Information Act — sourced from USCIS, obtained by Bloomberg — covering cap-subject petitions that were selected and filed in FY2021–FY2024.
Selection figures last revised August 10, 2026. Nothing on this page is legal advice.