Net migration has nearly halved. But fewer people are choosing to come — not fewer arriving by boat.

Net migration is at a four-year low — but gross immigration remains historically high

Net migration to the UK — people arriving to live for at least a year, minus those who leave — was provisionally estimated at 171,000 in the year to December 2025.¹ That is almost half the 331,000 recorded in December 2024, and 82% below the peak of 944,000 in the year to March 2023.¹

The 2025 level matches figures last seen when the post-Brexit immigration system launched in early 2021 during Covid travel restrictions. It is also, notably, below the net migration levels typical of the mid-2010s. However, the gross immigration figure — 813,000 people arriving — remains far above the 500,000–650,000 range that characterised the decade before the pandemic.⁷ That comparison should be read with care: the pre-2020 range comes from the old International Passenger Survey, while the 813,000 is built from administrative records that capture arrivals the survey missed — the new method has added several hundred thousand to cumulative counts over four years — so part of the apparent gap reflects better measurement, not only more people. Even so, net migration has fallen sharply while the scale of arrivals has not returned to pre-2020 norms.

The Office for National Statistics cautions that all current figures are "official statistics in development": the admin-data methodology, backdated to 2012, has been revised several times and remains subject to further updates.² Estimates from before June 2021 use a different method and are not directly comparable.

Stretching the horizon back three decades puts the 2025 figure in context. Net migration in the early 1990s ran near 44,000 a year; it climbed through the 2000s and 2010s, reached a recorded all-time high of 944,000 in the year to March 2023, and has fallen back sharply since.⁷ The longer view also makes the data’s central limitation visible: the figures before and after 2021 come from different measurement systems, so the series below is shown as discrete, individually sourced reference points rather than a continuous line.

Two policy decisions did most of the work: care workers and student dependants

The fall in arrivals is highly concentrated. Health and Care Worker visa grants fell by around 85% from their 2023 peak of approximately 156,000, to just 23,000 in 2024, after the previous Conservative government tightened sponsor requirements and barred most care workers from bringing dependants.⁵ Student dependant visas collapsed 83% in a single year after a January 2024 rule change restricted them to research-based postgraduate students only.⁵

The ONS reports that work-related non-EU immigration fell 47% in 2025 alone.¹ The Migration Advisory Committee, the independent body that advises government, attributed the earlier post-pandemic surge to three factors: a rise in work visas especially in health and social care, new humanitarian routes for Ukrainians and Hong Kong BNO holders, and increased overseas student recruitment as university finances deteriorated.⁴ The current Labour government added further restrictions in May 2025, including closing the care worker route to new overseas recruitment, with the Home Office estimating those changes would reduce visa grants by a further 98,000.³ Those restrictions are not yet fully visible in the 2025 data.

813,000 in, 642,000 out: who is moving and in which direction

The headline figure of 171,000 is the balance of two large and opposite flows. In 2025, 813,000 people made a long-term move to the UK, while 642,000 left for a year or more. Among nationality groups, only non-EU nationals show positive net migration; both EU nationals and British citizens are in net emigration, with more leaving than arriving.¹

The 278,000 non-EU nationals who left in 2025 are largely an echo of the 2022–23 surge: more than half were former students whose visas had expired.¹ British emigration, estimated at 246,000, appears high but the ONS cautions this reflects a methodological change in November 2025 rather than a new outflow trend — the new RAPID method captures departures that were previously unmeasured.²

The reversal is clearer still over a decade. In 2015, near the mid-2010s peak, EU nationals were the largest arriving group; today they are the smallest, while non-EU arrivals have more than doubled. Total immigration actually rose over the ten years — from roughly 632,000 to 813,000 — even as net migration fell, because emigration rose faster.⁷

Small boat crossings rose again in 2025. They are not what drove net migration down.

While net migration fell, arrivals by small boat across the English Channel moved in the opposite direction. A provisional 41,472 people crossed in 2025 — a 13% rise on 2024's 36,816, and the second-highest annual total since the Home Office began systematic recording in 2018, when 299 people made the crossing.⁸ Only 2022's 45,774 arrivals was higher.

Small boat arrivals are a legally and statistically separate measure from the ONS net migration figures. The key distinction: net migration counts people who move for at least 12 months; small boat crossings are counted on arrival regardless of what happens next. Virtually all who arrive by small boat (99% in 2025) subsequently claim asylum,⁹ and asylum applicants are included in the ONS figures only once they are assessed as long-term migrants. Small boat arrivals accounted for 41% of all UK asylum claims in 2025.⁹

The grant rate for asylum claims from small boat arrivals, measured across 2018–2025, was 62% — higher than the overall grant rate for asylum applicants.⁹ Returns have been limited: approximately 7,500 of the 193,000 people who arrived since 2018 had been removed from the UK by end-2025, a return rate of around 4%.⁹

Twenty-four people died attempting the crossing in 2025, down from 73 the previous year.⁹ The UK and French governments agreed a £662 million deal in April 2026 to fund enforcement operations on northern French beaches over three years.⁸ A "one-in, one-out" returns pilot with France ran from August 2025 until 11 June 2026; as of early March 2026 it had returned 377 people and accepted 380 in exchange — a near net-neutral outcome.⁸

The most recent data shows early signs of a slowdown: around 9,000 people crossed in the first five months of 2026, 38% fewer than in the same period of 2025, though the busier summer season has not yet fully passed.⁹

Study remains the largest single reason — asylum is 14%

Of the 627,000 non-EU nationals who arrived in 2025, 47% came for study-related reasons, 23% for work, 14% as asylum applicants, and 7% for family reunion.¹ These figures are based on the original visa type or asylum status on arrival and do not capture people who later switch visa categories. EU nationals, who made up 9% of total arrivals, are not broken down by reason in the same way because many hold EU Settled Status or residual rights.¹

Study has been the largest single reason for non-EU migration for most of the past two decades, and the recent fall has been concentrated in work rather than study. Non-EU work immigration peaked at roughly 471,000 in the year to mid-2023 before falling to about 146,000 on a reason basis in 2025, as the care-worker and dependant restrictions took effect; study volumes fell more modestly.¹ A fuller decade-by-reason series can be built from the Home Office’s published visa grants by category, which run back to 2005.⁵

Lower net migration: public mandate met, or workforce problem created?

The fall in net migration reflects stated government policy and responds to persistent public concern. Whether it is the right outcome involves both empirical questions (what are the workforce effects?) and value questions (how should society weigh population size against economic need?) that evidence alone cannot resolve.

Key data

What this piece left out

• Asylum backlog in full. The number of cases awaiting an initial decision fell to 35,744 (about 48,758 people) at the end of March 2026 — roughly 55% lower than a year earlier and the lowest since September 2019, after a concerted push to clear the backlog. The asylum grant rate at initial decision has also eased from its post-2022 highs (around 45% in the year to September 2025). The backlog’s composition and the rise in appeals are a substantial story in their own right.
• Regional impact. ONS does not publish local authority breakdowns from this source; the geographic distribution of both legal and irregular arrivals is not captured here.
• UK nationals abroad. Approximately 1.3 million UK nationals live in EU countries. The reasons why British people are leaving — and whether the new RAPID data represents a genuine trend shift — merits dedicated analysis.
• ONS methodology in transition. The move from the International Passenger Survey to admin-data (RAPID for British nationals, HOBI for non-EU and EU) is ongoing. The implications for data comparability across time deserve separate treatment.