The record library
Every extreme in seventy-five years of data — all mechanically computed, every row clickable into the map that proves it. Each line is a lead anyone can cite.
Cette page n'est pas encore disponible en français — la version anglaise est affichée ci-dessous.
Three things before reading: this page uses the "refugees" metric (the Venezuelan exodus is mostly under OIP — see the definitions timeline); pre-1967 definitions were narrower; and a decline is not necessarily returns — it can be naturalisation, revisions, or definition changes.
Largest one-year rises in hosting, ever
| # | Country | Year | Δ |
|---|---|---|---|
| 1 | Iran | 2022 | +2 626 748 |
| 2 | Iran | 1990 | +1 324 401 |
| 3 | Russie | 2022 | +1 266 771 |
| 4 | Iran | 1981 | +1 235 000 |
| 5 | Congo-Kinshasa | 1994 | +1 152 295 |
| 6 | Pakistan | 1980 | +1 028 230 |
| 7 | Pakistan | 2007 | +990 561 |
| 8 | Turquie | 2014 | +977 434 |
| 9 | Pologne | 2022 | +966 254 |
| 10 | Turquie | 2015 | +953 983 |
Largest one-year rises by origin, ever
| # | Country | Year | Δ |
|---|---|---|---|
| 1 | Ukraine | 2022 | +5 656 615 |
| 2 | Afghanistan | 2022 | +2 948 848 |
| 3 | Afghanistan | 1981 | +2 145 063 |
| 4 | Rwanda | 1994 | +1 807 111 |
| 5 | Syrie | 2013 | +1 739 312 |
| 6 | Syrie | 2014 | +1 419 167 |
| 7 | Afghanistan | 1980 | +1 234 921 |
| 8 | Irak | 2006 | +1 188 606 |
| 9 | Soudan du Sud | 2017 | +1 003 181 |
| 10 | Syrie | 2015 | +985 746 |
Largest one-year declines in hosting, ever
Again: a decline may be returns — or naturalisation, revisions, or definition changes. Click into the year and the country page for context.
| # | Country | Year | Δ |
|---|---|---|---|
| 1 | Iran | 2025 | -2 718 743 |
| 2 | Iran | 1993 | -1 655 685 |
| 3 | Pakistan | 1992 | -1 470 673 |
| 4 | Somalie | 1981 | -1 300 000 |
| 5 | Russie | 2024 | -1 220 054 |
| 6 | Pakistan | 2002 | -971 363 |
| 7 | Congo-Kinshasa | 1996 | -757 788 |
| 8 | Liban | 1978 | -695 000 |
| 9 | Malawi | 1994 | -623 363 |
| 10 | États-Unis | 2007 | -562 280 |
All-time per-capita hosting peaks
Refugees per 1,000 residents (each country at its own all-time peak year; population ≥ 100k).
| # | Country | Year | per 1,000 |
|---|---|---|---|
| 1 | R.A.S. chinoise de Hong Kong | 1964 | 357.1 |
| 2 | R.A.S. chinoise de Macao | 1964 | 354.5 |
| 3 | Somalie | 1980 | 336.9 |
| 4 | Liban | 1976 | 256.7 |
| 5 | Belize | 1988 | 171.8 |
| 6 | Djibouti | 1991 | 152.3 |
| 7 | Gabon | 1977 | 129.7 |
| 8 | Guinée équatoriale | 1968 | 125.5 |
| 9 | Malawi | 1992 | 104.4 |
| 10 | Monténégro | 2023 | 102.2 |
Longest-running #1 origin in the world
Number of years spent as the world’s largest origin of refugees, across 75 years.
| # | Country | Years |
|---|---|---|
| 1 | Afghanistan | 34 |
| 2 | Royaume-Uni | 29 |
| 3 | Syrie | 10 |
| 4 | Éthiopie | 1 |
| 5 | Ukraine | 1 |
"Unknown" (UNK) is listed as-is: in early decades UNHCR recorded large caseloads with no origin attached. That is an honest artifact of the historical record, not a data error — we do not guess or backfill.
All figures computed daily from the data (source: UNHCR Refugee Population Statistics Database, data as of 2025-12-31). Methods and limits: methodology; citing: cite responsibly.