๐Ÿ›‚ U.S. Department of State ยท Bureau of Consular Affairs ยท Nonimmigrant Visa Statistics travel.state.gov โ†—
Government Data Gap ยท Original Reconstruction ยท August 18, 2026

State Never Published FY2025 Visa Data. So We Rebuilt It.

Every year we come back to State's "Nonimmigrant Visas by Individual Class of Admission" table to see how many H-1Bs, H-2As, H-2Bs and other work visas the U.S. actually issued. This year the FY2025 table (fiscal year ended September 30, 2025) simply isn't there โ€” nearly eleven months after the year closed, well past the usual publishing lag. State does keep publishing monthly issuance PDFs on schedule, though, so we pulled and parsed all twelve of them ourselves and reconstructed the FY2025 numbers directly.

FY2025 Total: 10,707,238 Visas 12 Monthly PDFs Parsed 10 of 12 Months Exact Match Annual Report: Still Unpublished
FY2025 Total NIVs
10.71M
FY2024 Total (Official)
10.97M
H-1B, FY2025
214,596
H-2A, FY2025
318,317
H-2B, FY2025
139,736
Visa Classes Tracked
77
๐ŸŽฏ How this started: Every year we pull State's Table XV(B)/XVI(B) โ€” "Nonimmigrant Visas Issued by Classification" โ€” to update our visa data browser. This year FY2025's version of that table doesn't exist yet on travel.state.gov, even though the fiscal year closed September 30, 2025. State's own monthly "NIV Issuances by Nationality and Visa Class" reports, however, are current โ€” so we fetched all twelve, parsed each one directly from its raw PDF content streams, and summed them into the annual breakdown State hasn't gotten around to publishing.

Methodology

State publishes two different products from the same underlying consular database. The annual table (Table XV(B) in recent years, formerly XVI(B)) gives class-of-admission totals for a full fiscal year, but as of this writing it stops at FY2024. The monthly product โ€” "NIV Issuances by Nationality and Visa Class," a roughly 80-page PDF per month listing every (country, visa class) pair โ€” is current through September 2025. We fetched all twelve FY2025 monthly PDFs directly (October 2024 through September 2025) and parsed each one's raw PDF content streams โ€” decompressing the embedded text objects and reading off each row's nationality, visa class, and issuance count โ€” rather than transcribing ~1,000 pages by hand. Every month's reconstructed total was cross-checked against that same PDF's own printed grand total: 10 of 12 months matched exactly, and the other two (December and April) were off by 11 and 2 units respectively out of totals near 900,000 โ€” rounding noise, not a methodology problem. The one systematic quirk: Mexico's Border Crossing Card entries (B1/B2/BCC and B1/B2/BCV) render with a font-encoding glitch in State's own PDF generator that garbles the label text; the issuance counts themselves are unaffected and we relabeled them correctly before publishing.

Guest Worker Visa Classes โ€” FY2024 (Official) vs. FY2025 (Reconstructed)

ClassWhat It IsFY2024 (Official)FY2025 (Reconstructed)Change
H-1BSpecialty occupation worker219,659214,596-2.3%
H-1B1Chile/Singapore treaty specialty worker3,0703,009-2.0%
H-2ATemporary agricultural worker315,328318,317+0.9%
H-2BTemporary non-agricultural worker139,541139,736+0.1%
H-3Trainee924533-42.3%
H-4Dependent of H visa holder139,874146,158+4.5%
L-1Intracompany transferee71,79977,247+7.6%
L-2Dependent of L-1 holder73,31777,029+5.1%
TNUSMCA professional (Canada/Mexico)15,67216,877+7.7%
E-3Australia treaty specialty worker3,9333,758-4.4%
E-3DDependent of E-3 holder3,5743,463-3.1%
E-3RE-3 visa renewal3,1002,812-9.3%
CW-1CNMI transitional worker2,5171,110-55.9%
CW-2Dependent of CW-1 holder651268-58.8%
All NIV classes, total10,969,93610,707,238-2.4%

FY2024 figures are State's own published Table XV(B) numbers. FY2025 figures are this reconstruction. The CW-1/CW-2 drop tracks the CNMI transitional-worker program's statutory phase-down, not a parsing artifact.

Now Live in Our Visa Data Browser

We loaded the full 77-class FY2025 breakdown into visa_browser.php alongside every fiscal year back to 1987 โ€” annual totals, per-class trends, a full year-by-year breakdown, and search. It's the same tool we've published every year; FY2025 just arrived a different way this time.

ExploreWhat you'll find
Annual TotalsAll 39 fiscal years, 1987-2025, worker/dependent/tricky category breakdown, visas-per-job vs. BLS payroll data
By Visa ClassTrend line for any single class (H1B, H2A, TN, etc.) across every year we have data for
FY2025 Full BreakdownAll 77 classes reconstructed for FY2025, ranked by volume, with category badges and % of total
SearchFind any class by code or description across all years

Investigative Assessment

Watch Nearly eleven months after FY2025 closed, State still hasn't published the annual class-of-admission table โ€” well outside the usual few-month lag between prior fiscal year-ends and their Table XV(B) releases. We don't know the specific cause (State-side staffing, the late-2025 government funding lapse, or something else), but the monthly product kept publishing on schedule throughout, which is what made this reconstruction possible at all.
Watch H-3 (trainee) issuances fell 42% year over year, and CW-1/CW-2 (CNMI transitional workers) fell more than 55% โ€” both small-volume categories where a modest absolute change produces a large percentage swing. CW-1/CW-2's decline lines up with that program's statutory phase-down schedule; H-3's drop is worth watching for another year before reading a trend into it.
Positive The two headline guest-worker categories most tied to labor-market displacement โ€” H-1B and H-2B โ€” barely moved (-2.3% and +0.1%). H-2A, agricultural labor, ticked up less than 1%. None of the major work-visa categories show the kind of surge that would itself be the story this year; the story is that State made us go find the numbers ourselves.

Help Support This Research

This investigation is produced independently โ€” one person, 23 years of data work, living on $1,419/month Social Security. If this work matters to you, a small donation makes a real difference.

♥ Donate via Square More Reports

💬 Comments