0.52.5BERT/dayNo comparable data★ Hotspot (a place, not an average)
1 BERT = 8.2 microsieverts = one day of the world-average natural dose.
An honest map. The figures come from different sources and different years, so treat them as approximate. As a rule of thumb the estimates carry roughly ±0.17 BERT/day of slack (about ±0.5 mSv/yr), so any two countries closer than that are a tie, not a ranking.
Sources and data notes
Converting radon to dose. Where a country published a radon concentration but not a dose, we used the UNSCEAR convention - dose (mSv/yr) ≈ concentration (Bq/m³) × 0.025 - which bundles an equilibrium factor of 0.4, 80% indoor occupancy and a dose coefficient of 9 nSv per Bq h m⁻³. The newer ICRP coefficient is about twice that, so some national figures elsewhere will look correspondingly higher.
UNSCEAR (world avg revised to ~3.0 mSv/yr, UNSCEAR 2024 Report Annex B, announced 2026); CNSC (US 3.0, Canada 1.8, Yangjiang 6.3, Kerala 12.5); Wikipedia Background radiation (UK <2); STUK (Finland 5.1, radon 4.0); KINS 2020 (South Korea 5.24); IRSN/INRS (France 3.0); Foro Nuclear (Spain 2.4-3.0); Min-Saúde/APA (Portugal 2.4); BfS (Germany 2.1); ARPANSA (Australia 1.7); St. Petersburg Inst. of Radiation Hygiene (Russia 3.3); SSM (Sweden 3.0); DSA (Norway 4.1); LIFE-Respire/PMC (Italy 3.5); CLOR / Nofer Institute (Poland: national mean 49.1 Bq/m³, published indoor-radon dose 1.36 mSv/yr, regional means 25-123 Bq/m³; EC-JRC puts Polish terrestrial gamma among Europe's lowest at ~0.3 mSv/yr); SURO, Czech national radon programme (country average 140 Bq/m³ ≈ 3.5 mSv/yr radon; EC-JRC puts radon above 75% of the Czech natural dose); Croatian national survey of 782 randomly sampled dwellings (68 Bq/m³ arithmetic mean = 2.2 mSv/yr radon, county means 33-198); Nuclear Safety Research Association / NIRS / JCAC nationwide surveys (Japan 2.1: cosmic 0.29, terrestrial 0.33, radon and thoron 0.59, ingestion 0.99 - every component measured domestically; prefecture range 0.72-1.24 mSv/yr for the external plus radon part); Mexican state-by-state indoor-radon means, UNAM and ININ (most states 60-100 Bq/m³ ≈ ~2 mSv/yr radon, full spread 51-1,863, highest state dose 3.13 mSv/yr in Chihuahua); ARN national indoor-radon survey, 1,994 dwellings across geologic zones (Argentina: 33.8 Bq/m³ country average = 0.84 mSv/yr radon, plus ARN figures of 0.39 cosmic and 0.46 terrestrial); South African radon studies (non-mining control areas 1.2-1.6 mSv/yr, plus ~0.45 cosmic from the Highveld altitude; Witwatersrand tailings dwellings up to ~16); high-background-area review (PMC) for Guarapari and Mamuju; IAEA/INIS Natural background radiation and population dose in China (China 2.3); Sohrabi et al., national survey of 1,000 houses in 36 cities (Iran: terrestrial plus cosmic 0.86 mSv/yr, of which 0.39 terrestrial from the national soil survey, leaving ~0.47 cosmic - high because much of Iran sits above 1,000 m) with Iranian indoor-radon studies averaging 2.6 ± 2.4 mSv/yr; Upper northern Thailand radon study, 192 houses across 8 provinces (average dose 4.27 mSv/yr, provincial means 75-168 Bq/m³ against a national average of 16-36; a quarter of regional lung-cancer deaths attributed to radon); Greenland indoor-radon survey, 459 year-long measurements in 257 homes (Narsaq 139 Bq/m³ = 6.5 mSv/yr radon, against 10.5 Bq/m³ in Nuuk - Narsaq sits ~7 km from the Kvanefjeld uranium deposit); Ramsar HBRA review, PMC (inhabitants receive up to ~80 mSv/yr; 260 is the peak reading at uninhabited hotspots); ATSDR Toxicological Profile for Ionizing Radiation (Denver ~6). Brazil and India are left grey. Both have real national measurements - a cosmic map and a country-wide terrestrial survey respectively - but neither has measured radon, which is around 60% of the dose, so any national total would be mostly assumption.