How the United States granted its green cards through adjustment of status, category by category, between fiscal years 2017 and 2026 — read from the official USCIS data.
The I-485 splits across six legal categories. A single one — Family — holds half the volume, and where the cases are most numerous is also where the most are denied.
Annual volume received, by base. Family grows nonstop, Employment had a single spike — and Cuba appears out of nowhere at the end.
For six years, Cuba was a minor line — between 17,000 and 60,000 applications per year.
Then came 2023.
In four years, the program grew tenfold — and one in four green cards through adjustment became Cuban.
But cases arrive faster than USCIS can decide them. The Cuban bottleneck grew 17× to surpass 355,000 pending cases — the second largest in the system, after Family.
Among cases that reach a decision, the humanitarian categories hover around 95-96%. The gap between the most permissive base and the strictest is less than 10 points.
Applications pending as of May 2026. Family and Cuba — two bases — hold 61% of the queue.
The first monthly national reading from the May 2026 snapshot: the queue keeps growing because far more applications come in than get decided.
At May's pace, the national humanitarian queue (628,698 cases) would take ~652 months to clear. It's the same brake signal seen office by office in the next chapter.
A single month is noisier than a quarter; "months to clear" is a descriptive metric of the national pace, not a personal wait-time prediction.
At the close of FY2026 Q1, the 1.33M pending then were broken out across 90 field offices and 6 service centers — and over time: 49 comparable quarters, FY2014 → FY2026. This geographic breakdown is the snapshot of that quarter; the national total is already at 1.57M (see the May flash, above). Source: DHS / USCIS Office of Performance and Quality.
Months to clear the queue at the current decision pace, by category:
"Months to clear the queue" is a descriptive metric of each office's pace (pending ÷ monthly decisions in the quarter). It is NOT a personal wait-time prediction: it does not account for each case's age or category prioritization.
Florida's humanitarian queue, quarter by quarter since FY2014. For nine years it was a boring story. Then it stopped being one — in four acts.
For 9 years the queue swung between 7,893 and 39,781 and always self-corrected: the FY2018 Q1 peak was worked down to its low in FY2020 Q2.
The wave of Cuban Adjustment eligibility after CHNV parole tripled receipts in a single quarter: from 12,293 to 43,057 (FY2023 Q3). Pending: from 33,392 (FY2023 Q2) to 108,309 (FY2024 Q1).
USCIS scaled decisions to record highs — 34,139 in a single quarter (FY2024 Q3) — and the queue FELL from 111,213 to 79,876: the system was catching up to the wave.
Decisions collapsed ×52.6 from the peak, while 13K–40K cases kept arriving each quarter. Pending: from 79,876 to 178,269.
In FY2024 Q3 it decided 34,139 cases in one quarter, in Florida alone. Six quarters later it decides 649.
In the top 5, the dominant category is humanitarian.
The overall 89.8% blends a large, strict base (Family, ~87%) with small, nearly automatic bases (Asylum, Refugee, Cuba: 95-96%). Reporting a single average number hides two distinct systems.
10× growth since 2023, today one in four I-485s, with an approval rate that held at 95%+ — and a bottleneck that grew 17× to 355K cases.
292K applications in one year resulted from unused family visa numbers that rolled over to the employment category during COVID. Worth remembering when modeling the future: Employment lives in a band of ~110-130K/year.
Family (605K) by structure — it was always the largest. Cuba (355K) by speed — cases arrived faster than decisions. Their combined share fell from 65% to 61%, not because they improved, but because Employment (+56K) and Refugee (+38K) grew faster in 2026. Any wait-time analysis starts here.
95-98% approval, almost never denied — because eligibility was already established at the prior stage (the grant of asylum or refugee status). The I-485 here is more administrative paperwork than adjudication.
Geographic chapter: FY2026 Q1 per-office dataset (DHS / USCIS Office of Performance and Quality, NPD, query 01/2026, PAER0020062), 99 rows reconciled programmatically against the official total and against the main report's figures. Offices with no activity excluded from the selector: Chula Vista (CA), Dover (DE), and Charleston (WV). Time series: FY2014 Q1 – FY2026 Q1, 49 comparable quarters normalized from 95 official files; FY2015 Q1 has no per-office breakdown (marked discontinuity); FY2013 Q4 excluded due to overlap with legacy cumulative reports.
231 I-485 rows from the USCIS master file, classified by legal base according to the form title (Cuban / Indochinese / Refugee / Asylum / Employment / Family / Other). Six bases with full coverage across 36 quarters; Indochinese, discontinued after FY2019 Q1, excluded from rate calculations. Approval rate = approved ÷ (approved + denied). Partial years (FY2017 = Q2-Q4; FY2026 = Q1) marked throughout the report. All figures computed programmatically; none estimated.
May 2026 snapshot: Report to Congress, Number of Service-Wide Forms by Month (Consolidated Appropriations Act 2022, P.L. 117-103), ELIS/CLAIMS3, queried June 2026. A different reporting stream from the quarterly series; counts may differ marginally between streams. Pending figures were updated to this snapshot; the historical series and the geographic breakdown remain at FY2026 Q1.
inmigreat
© 2026 · I-485 report by adjustment base
Official USCIS data · This report does not constitute legal advice