Net migration
ONS Long-Term International Migration provisional estimates, year ending December 2025. Released 21 May 2026. The headline figure has fallen 82 percent from its YE March 2023 peak, but methodology change accounts for a substantial part of that move and the figure will be revised.
95 percent uncertainty interval: 145,000 to 197,000.
ONS notes the interval does not include uncertainty from the emigration re-arrival adjustment, so the real spread is wider.
Net migration trend (last 16 periods)
YE Mar 22 575,000
YE Jun 22 681,000
YE Sep 22 822,000
YE Dec 22 891,000
YE Mar 23 944,000
YE Jun 23 924,000
YE Sep 23 891,000
YE Dec 23 848,000
YE Mar 24 734,000
YE Jun 24 649,000
YE Sep 24 R 497,000
YE Dec 24 R 331,000
YE Mar 25 P R 260,000
YE Jun 25 P R 219,000
YE Sep 25 P 202,000
YE Dec 25 P 171,000
Period codes: P = provisional, R = revised, no suffix = completed. The peak of 944,000 (YE Mar 2023) is itself a revised figure; previously published as 906,000.
Non-EU+ immigration by reason, YE Dec 2025
Study-related 14k Main: 14k · Dep: 0
Work-related 16k Main: 12k · Dep: 4k
Asylum 0 broadly stable
Family 3k
Other 1k
Work-related immigration fell 47 percent year on year. The Home Office summary attributes most of this to the Health and Care Worker visa restrictions, with Caring Personal Service grants falling from 108,000 to 1,400.
Revisions tracker
Each provisional estimate is revised twice before settling. The May 2026 vintage of YE December 2024 is the third (completed) estimate. Migration Observatory: this revision results from a change of methodology, not a change in the underlying trend.
PeriodPublicationStageImmigrationEmigrationNet migration
YE Dec 24 May 25 Initial provisional estimate 948,000 517,000 431,000
YE Dec 24 Nov 25 Revised provisional estimate 1,015,000 669,000 345,000
YE Dec 24 May 26 Completed estimate 1,012,000 680,000 331,000
YE Dec 24 May 26 Revision for 12 months travel data -3,000 11,000 -14,000
PeriodPublicationStageImmigrationEmigrationNet migration
YE Sep 24 May 25 Initial provisional estimate 1,104,000 503,000 601,000
YE Sep 24 Nov 25 Revised provisional estimate 1,168,000 661,000 508,000
YE Sep 24 May 26 Completed estimate 1,159,000 662,000 497,000
YE Sep 24 May 26 Revision for 12 months travel data -9,000 1,000 -11,000
ONS vs Home Office vs DWP triangulation
The three official sources do not reconcile by design. The Home Office counts visa grants. DWP counts NINo allocations after arrival. ONS LTIM counts long-term arrivals. The same person can be in all three or none.
Period Non-EU+ LTIM Non-EU+ visas granted (HO) Non-EU+ NINo (DWP)
YE Mar 25 P R 622,000 725,000 583,000
YE Jun 25 P R 577,000 702,000 571,000
YE Sep 25 P 557,000 692,000 586,000
YE Dec 25 P 533,000 664,000 550,000
Home Office figures exceed LTIM because not every visa grant leads to a long-term stay. DWP NINo allocations capture people registering to work, which excludes dependants and short stays.
What is not counted
Visa overstayers without asylum claims
Assumed to have emigrated under the current ONS method. If overstaying has risen, net migration is understated.
Irregular migrants who do not claim asylum
Never counted. The small boat arrivals figure (39,000 YE Mar 2026) captures only those who claim asylum after arrival.
Pre-June 2021 comparability
The IPS-based estimates before June 2021 are NOT directly comparable to the current admin-based RAPID and HOBID figures. British emigration YE Dec 2024 alone moved 77,000 to 257,000 with the methodology change.
EU Settlement Scheme grants
371,000 settled-status grants in YE Mar 2026 do not count as immigration in LTIM because they confirm status for people already in the UK. They affect population denominators.
Methodology notes
ONS replaced the International Passenger Survey with the Home Office Borders and Immigration Data (HOBID) for EU and non-EU visa holders, and the Department for Work and Pensions Registration and Population Interaction Database (RAPID) for British nationals. The shift was applied retroactively from year ending June 2021 onwards. Estimates before that period use the older IPS method and are not directly comparable.
Source: Long-term international migration, provisional: year ending December 2025. Companion ONS article: UK emigration explained. Independent commentary: Migration Observatory.