Every I-130 petition travels the same path: it waits for a USCIS decision and — if approved — waits for a visa number. We follow that journey with official USCIS data, stop by stop. Three reports, one single story.
None of these figures is the same queue. They are different stops on the same path — and we're going to walk them in order.
Pending is 67% preference. But the real backlog is 91% immediate relative. It's an inverted mirror — and there's a reason.
Why: USCIS defines "backlog" as what's pending beyond its target time and within its control. Almost all preference pending is excluded — those people aren't waiting on USCIS, they're waiting on a visa number.
The backlog measures how slow USCIS is. The visa queue measures the limits set by Congress. They are not the same problem.
The I-130 is adjudicated at service centers, not at the field offices where you have your interview. Five centers hold 87.6% of all pending.
It's the office you go to in person: your biometrics appointment and, if your case requires one, your interview. It serves the public and handles local paperwork — but it decides almost no I-130 petitions.
These are centers closed to the public where forms sent by mail are processed and decided. Your I-130 is adjudicated here. Five centers hold almost all of the country's workload.
Its queue fell 37% (−210K), the biggest contribution to the total decline. "Rest of the world" had the most dramatic peak-and-recovery: it rose to 862K and came back down to 510K.
Spouses and minor children of residents are the highest-priority category. Visas are almost always available, so no one waits — the queue reads zero.
When demand exceeds the cap, a brief queue forms (the 380K peak in 2023) that drains back to zero. Its numbers are episodic, not a trend.
Country × category, FY 2026 Q1. Color marks the intensity of the queue. Hover over each cell to see the detail.
2.3M under review + 1.03M in the visa queue are not 3.3M people. They are different stages of the same path; adding them is double-counting.
One petitioner can file several, and derivative beneficiaries — spouse, children — aren't counted separately. The real number of people is higher.
Except for quarterly received/approved/denied, everything is a point-in-time stock as of Dec 31, 2025 — a snapshot, not a count of arrivals.
A falling queue means it's emptying — visas issued faster than petitions coming in — not that fewer people are migrating. The −21% is good news for those who wait.
From September 2023 to today, nine consecutive quarters of decline, down to 1.03M — the best level in five years.
Reforming the family visa queue is, mathematically, reforming the siblings queue.
−37% (−210K), more than India, China and the Philippines combined. The center of gravity of family visa demand.
The longest-waiting category — spouses and children advance quickly; siblings wait decades.
Spouses and children of residents barely wait: the queue is usually zero. Its peaks are episodes of visa availability, not a backlog.
The backlog USCIS can resolve and the one that depends on Congress are opposite problems. Confusing them is the most common mistake when reading this data.
The wait is decided by USCIS logistics — Vermont, California, Texas, Nebraska, Potomac — not by your zip code.
Three official USCIS reports (Office of Performance and Quality, accessed 01/2026), combined into a single journey: performance and pending by office (PAER0020061, Oct–Dec 2025), Net Backlog by form type (as of Dec 31, 2025) and approved petitions waiting for a visa by category and country (PAER0020057, 17 snapshots FY2021 Q3→FY2026 Q1). All values are point-in-time counts (stock), except quarterly received/approved/denied (flow). The aggregates reconcile to the figure (office→total, country→TOTAL, category→ALL). The frontlog is published only at the all-forms level, not per I-130, which is why it's left out.
inmigreat
© 2026 · I-130 Report · The complete journey of a family petition
Official USCIS data · This report does not constitute legal advice