This guide was built from government advisories (US State Dept, UK FCDO, Australian Smartraveller), official statistics where they exist, and local news sources. See methodology →
Quick Safety Summary
What Nobody Tells You
1. What Changed: Canada and Montreal on Top, the Netherlands Out of the Top Five
Canada is the safest country in the world for 2027 and Montreal the safest city, according to Berkshire Hathaway Travel Protection's 12th annual State of Travel Insurance Safest Destinations report, released on 28 October 2026. The country top three is Canada, New Zealand and Australia; the city top three is Montreal, Zurich and Amsterdam. The rankings blend a survey of 2,784 travelers, fielded by Polymath Research + Marketing from 14 to 28 July 2026, with external sources including the World Health Organization, the Global Peace Index, Numbeo, the Equaldex Equality Index, the Global Terrorism Index, State Department ratings and GeoSure Global. The moves, per the release: Canada climbed from fifth to first, New Zealand from sixth to second, and Australia slipped to third after holding second place for the past two years. Japan jumped from ninth to fifth. Among cities, Montreal made what BHTP calls a debut at No. 1, Zurich rose from third to second, Amsterdam from fourth to third, and Reykjavik fell from first to fifth. The headline faller is the Netherlands — last year's safest country. It sits eighth in the release's numbered list (and on BHTP's own country page), although one sentence of the release's prose says it "fell from first to ninth". Either way, the 2027 report reverses most of the 2026 one: a year ago the Netherlands had jumped from fourteenth to first. Single-year swings of that size are a known property of survey-driven rankings, and this year supplies the second consecutive example. Set against the 11th edition's list, other moves stand out too: the United Arab Emirates, seventh last year, is absent from the 2027 top fifteen, while Norway, Italy, Spain and Germany have entered it. None of this means Canada became safer, or the Netherlands more dangerous, between two summers. It means a survey-weighted composite registered a change in how a few thousand travelers rated places they had recently been. That is a legitimate thing to measure — and it is a different thing from what our own City Safety Index 2026 measures, which is why the comparison below is useful rather than a contest.
2. Where the Survey and the Arithmetic Agree
We cannot compare countries directly — our index scores 50 cities, not nations — but we can test BHTP's city list, and the overlap is striking. Nine of BHTP's fifteen cities are in our 50-city index; six are not: Montreal, Reykjavik, Oslo, Honolulu, Dublin and San Francisco. Of the nine we can score, six sit in our top ten: Zurich and Singapore (joint 1st, 1.4), Copenhagen (3rd, 1.6), Sydney (5th, 2.1), Amsterdam (9th, 2.4) and Tokyo (joint 10th, 2.5). The headline agreement is Zurich. Zurich, second on BHTP's city list, is joint first in ours at 1.4, level with Singapore. The two instruments share almost nothing in construction — one is a survey of where American travelers felt safe, blended with outside indices under unpublished weights; the other is seven analyst-assigned sub-scores under published weights — and they land on the same city within one place of the top. When unrelated methods agree, that is about the strongest signal either can give. The country list tells the same story at lower resolution. Seven of BHTP's top ten countries contain a city we score, and six of the seven BHTP top-ten countries we can test have their best-scoring city inside our top ten; the exception is Portugal, where Lisbon is joint 15th at 3.0. The other three — New Zealand, Iceland and Ireland — have no city in our index, so we say nothing about them. Canada deserves its own line, because it is BHTP's new number one and its new number-one city is one we do not cover. Montreal is not in our 50-city index, so we have no score for it and will not invent one. Toronto, Canada's only city in our index, is 8th at 2.3 — a top-ten result that is consistent with BHTP's verdict on the country, but it is Toronto, not Montreal, and the two should not be swapped in a headline. For the full top of our table, narrated city by city, see the safest cities in the world 2026.
3. Where They Diverge — and Why
Three cities on BHTP's top-fifteen list sit well down ours: Venice (joint 15th, 3.0), Barcelona and Paris (both joint 28th, 3.9). This is the most instructive disagreement in the comparison, and it has a specific cause. Venice's scam sub-score is a 5, Barcelona's street-crime sub-score a 6 and Paris's scam sub-score a 6 — the kind of tourist-facing petty crime and overcharging that our formula weights at 45% combined (crime 25%, scams 20%). A traveler survey sees these cities differently, and not wrongly. Pickpocketing is common in aggregate and rare per visitor: most people who spend four days in Paris come home with their phone, report a safe trip, and rate the city accordingly. A survey of firsthand experience averages over those trips. A sub-score asks a different question — how likely is the thing, and how concentrated is it where visitors stand — and records the answer even when the typical respondent never met it. The disagreement runs the other way too. Five of our top eleven are absent from BHTP's city list: Vienna (4th, 1.8), Seoul and Melbourne (joint 6th, 2.2), Toronto (8th, 2.3) and Osaka (joint 10th, 2.5). BHTP's list features cities American travelers visit in numbers — Honolulu, San Francisco, Montreal — and a top-fifteen cut cannot include everyone. A city missing from a survey-based list may simply be under-visited by the people surveyed, not rated poorly by them. And there is a resolution effect inside BHTP's own two lists. Switzerland is 12th on its country list, yet Zurich is its second-safest city. That is not a contradiction; a country score averages over everywhere a traveler might go, a city score over one place. It is exactly why we publish at city level and keep the seven sub-scores visible — so a reader can see which part of a place's record is doing the work.
4. What a Country Ranking Cannot See
Last year's report made the Netherlands the safest country in the world; this year's moves it to eighth. Both editions share a limitation that no amount of re-weighting fixes: a country ranking cannot see where in the country the risk sits. Our worked example is Amsterdam. BHTP puts it third among cities this year. Our index broadly agrees — Amsterdam is 9th at 2.4 — but it also carries a crime sub-score of 4, the worst in our top ten. The reason is concentration: in the first four months of 2026, Amsterdam absorbed 38.8% of all Dutch pickpocketing reports on roughly 5% of the country's population, almost all of it along one tram corridor and around Centraal Station, as we documented in our Amsterdam pickpocketing deep-dive. Both readings are true at once. The Netherlands is a very safe country by violent crime, health care and stability, and Amsterdam is a safe city with a well-mapped theft problem. A survey of how trips went will mostly hear the first fact; a sub-score has to record the second. The same logic explains the Gulf: BHTP's 11th edition placed the UAE seventh among countries, and it is absent from the 2027 top fifteen, while our index has carried Dubai at joint 43rd all year once political stability and the active US advisory are counted — a gap we unpack in our comparison of five safety rankings.
5. How to Use BHTP and Our Index Together
Use BHTP for the shortlist. Its survey counts only travelers rating places they actually experienced, its composite folds in health, terrorism and equality data, and as a reading of where American travelers feel confident it is well built. It is a good first filter for countries. Use a sub-score index for the city. Once you have a country, a ranking you can recompute tells you which of its cities carries which risk — Paris's scam score, Barcelona's street crime, Tokyo's earthquake exposure. Our City Safety Index 2026 publishes all 350 sub-scores and a CSV, so you can re-weight it toward what worries you. Use the deep-dives for the street. Neither instrument can tell you to keep your phone out of your back pocket on one Amsterdam tram line; a corridor-level investigation can. And when the two disagree, read the disagreement instead of picking a side. A city that BHTP's travelers loved and our sub-scores flag is usually a city with a petty-crime problem that rarely ruins a trip — until it ruins yours. That is the most practically useful sentence either ranking can produce — and it only appears when you hold both.
Final Verdict
Frequently Asked Questions
Canada. Berkshire Hathaway Travel Protection's 12th annual Safest Destinations report, released 28 October 2026, ranks Canada first, up from fifth, followed by New Zealand and Australia. The Netherlands, last year's number one, fell to eighth on the published list. The caveat: BHTP ranks countries, and a country rank says little about the specific city or district you will visit.
Montreal, Zurich and Amsterdam lead BHTP's 2027 city list, followed by Tokyo and Reykjavik, which fell from first to fifth. Our own index agrees on Zurich, which is joint first in ours at 1.4, and broadly on Amsterdam (9th) and Tokyo (joint 10th). Montreal and Reykjavik are not among the 50 cities we score, so we cannot confirm or contest them.
BHTP surveys travelers about their firsthand travel experiences — 2,784 respondents this year, fielded 14–28 July 2026 — and combines the results with external indices, including WHO data, the Global Peace Index, Numbeo, the Equaldex Equality Index, the Global Terrorism Index, State Department ratings and GeoSure Global, into a weighted composite. Past rankings are also factored in. The weights and per-destination scores are not published, so the ranking cannot be recomputed from outside.
At the top, largely yes. Of the nine BHTP top-fifteen cities that are also in our 50-city index, six sit in our top ten, and Zurich ranks near the top of both. The clearest disagreements are Venice, Barcelona and Paris, which make BHTP's top fifteen but sit mid-table in ours because of high scam and street-crime sub-scores. Our comparison of five major safety rankings explains why each crowns a different winner.
BHTP does not publish the scores behind the ranking, so the precise cause is not visible from outside. What the record shows is volatility: the Netherlands jumped from 14th to 1st in the 11th edition and fell back to 8th in the 12th. Large single-year moves are common in survey-weighted rankings and usually reflect shifts in respondent sentiment more than changes on the ground.