Brexit and immigration
The category Britain voted to control fell below zero. The category it already controlled grew.
Year ending December 2025
171k
Net migration
Down from 331k the previous year
813k
Immigration
People arriving long-term
642k
Emigration
People leaving long-term
−42k
EU+ net migration
Negative. More EU citizens leaving than arriving.
Handle these numbers carefully. These are official statistics in development, not accredited official statistics, and the ONS has revised them materially and repeatedly as its methods improved. Any single year’s figure should be read as an estimate with a wide margin, not a count.
The story is composition, not level
“Did Brexit reduce immigration?” is the wrong question, and both sides ask it because both have an answer they like. What the data shows is a change of shape.
| Nationality group | Immigration | Emigration | Net |
|---|---|---|---|
| Non-EU+ nationalsNow 77% of all immigration. Study remains the largest single reason for arrival at 47% of non-EU+ arrivals; work-related arrivals fell 47% in 2025. | 627k | 278k | +350k |
| EU+ nationalsNegative. More EU citizens are leaving than arriving — the direct, measurable consequence of ending free movement. | 76k | 118k | -42k |
| British nationalsAlso negative, and has been for years. British emigration is a large and consistently under-discussed component of the net figure. | 110k | 246k | -136k |
Thousands, year ending December 2025. ONS, Long-term international migration, provisional: year ending December 2025 ↗ — Official statistics in development — provisional, and subject to substantial revision.
Of 813,000 arrivals in the year to December 2025, 627,000 were non-EU+ nationals and 76,000 were EU+ nationals. Before 2021 those two groups were roughly comparable in size. The single most consequential migration statistic of the Brexit period is not the net total — it is that ratio.
How EU citizens arrive
BeforeFree movement. No visa, no cap, no route requirement, no salary threshold.
AfterThe same points-based system as everyone else — a job offer at a required skill level and salary, or a study place, or a family route.
Net EU migration
BeforeConsistently positive, and roughly half of total net migration at the time of the referendum.
AfterNegative: −42,000 in the year to December 2025.
Non-EU share of arrivals
BeforeRoughly comparable with EU arrivals.
After77% of all immigration — 627,000 of 813,000 arrivals.
Who decides the rules
BeforeFree movement was a treaty obligation. The UK could not restrict it unilaterally.
AfterEvery route is set in UK law and can be changed by a UK government without asking anyone.
This is the part of the promise that was unambiguously delivered.
Long-term international migration estimates. ONS, Long-term international migration, provisional: year ending December 2025 ↗
What happened between the referendum and now
- Year ending June 2016Referendum
Net migration around 330,000 — the figure that dominated the campaign, roughly half of it EU citizens exercising free movement.
- 2021New system begins
Free movement ends. The points-based system applies the same rules to EU and non-EU citizens, and is more open to skilled workers from outside Europe than the previous regime.
- Year ending June 2023Record peak
Net migration peaks at roughly 900,000 — around three times the level at the referendum — driven by non-EU work and study visas, plus Ukraine and Hong Kong schemes.
- Year ending December 2024Falling
Net migration 331,000 as visa rules tighten and the post-pandemic student surge unwinds.
- Year ending December 2025Latest
Net migration 171,000 — around half the previous year, and below the level at the time of the referendum.
Do not read this as one trend. These points span a pandemic, two humanitarian visa schemes, a student boom and bust, and several changes to ONS methodology. Reading them as a single Brexit trend line would attribute to one policy the combined effect of at least five different things.
Did Britain take back control of immigration?
Legally
Yes, without qualification. Free movement ended on 31 December 2020. Every route into the UK is now set by UK law and can be changed by a UK government without asking anyone.
Numerically
Not in the way most Leave voters expected. In the four years after control was regained, net migration rose to roughly three times its referendum-era level before falling back below it in 2025.
Both are true because control and reduction were always different things. The mechanism worked exactly as designed and was then used to admit more people, largely to staff the NHS and social care and to recruit international students. The people who promised control delivered it; the people who heard "fewer" heard something that was never in the mechanism.
The labour-market side
The end of free movement removed a flexible labour supply that several sectors had come to rely on — agriculture and food processing, hospitality, road haulage and social care most visibly. Some of those shortages were filled by non-EU recruitment, some by higher wages, and some not at all. Attributing any specific shortage to Brexit alone is difficult, because the pandemic disrupted the same sectors at the same time.
More: the full immigration section covers routes, asylum, small boats and countries of origin, and what immigration contributes models the fiscal effect.