What is a Travel Risk Score?

The short version

Our Travel Risk Score is a weighted average of twelve category scores, each rated 0–100, with weights chosen so that the categories most likely to actually disrupt a trip carry the most influence. The result rolls up into one of four tiers:

• Low (0–24) — the destination poses normal travel risks; standard precautions apply.
• Moderate (25–49) — one or two categories deserve attention; informed planning recommended.
• Elevated (50–69) — multiple risk factors compound; travelers should reconsider non-essential trips or take meaningful precautions.
• Extreme (70–100) — significant ongoing risks across multiple categories; travel often advised against by official sources.

You can run a free interactive assessment for any city to see its current score with the underlying category breakdown and live data sources.

How the 12 categories are weighted

The weights below were chosen after reviewing how government travel advisories, traveler insurance assessments, and security firms weight different categories. They sum to 1.0.

Category	Weight	What it captures
Crime	0.18	Petty theft, violent crime, scams targeting tourists, neighborhood-level variation
Health	0.15	Disease outbreaks, water quality, food-borne illness risk, mosquito-borne disease season
Transport & route	0.13	Road-traffic injury rate, public-transport reliability, airline safety record, route-specific hazards
Government advisory	0.12	Current US State Department advisory tier and equivalent country foreign-ministry warnings
Political & civil unrest	0.09	Protests, elections, ethnic or sectarian tension, terrorism, border instability
Natural disaster & weather	0.08	Earthquake/cyclone/flood/volcano probability, current weather extremes, monsoon timing
Personal modifier	0.06	Traveler-supplied factors: solo female, LGBTQ+, journalism, medical conditions, etc.
Environmental	0.05	Air quality, water contamination, heat index, altitude
Accommodation	0.04	Hotel safety standards, fire-code compliance, neighborhood when arriving after dark
Arrival time	0.04	Risk multipliers for late-night airport arrivals or arriving in unfamiliar areas after dark
Digital security	0.03	ATM skimming, public Wi-Fi risk, surveillance, mobile network reliability
Money safety	0.03	Counterfeit currency, taxi-meter scams, exchange-rate fraud, card cloning

See our full methodology page for the per-category scoring rubric.

Why a composite score is useful

The honest answer to "is X destination safe?" is almost never a simple yes or no. A city like Mexico City might be relatively safe for petty crime in tourist zones but elevated for road traffic and air quality. Tokyo might be one of the lowest-crime cities in the world but sits on a major earthquake fault. London is generally safe but has elevated digital-security risks for travelers using public Wi-Fi without a VPN.

A composite score is useful because it forces all of these signals into a single comparable number. You can answer questions like:

• Is destination A meaningfully riskier than destination B for this trip?
• What's the single biggest risk factor in this city — the thing I should prepare for most?
• How does this city's score change for my trip specifically (solo female traveler, business traveler with valuables, family with kids)?
• Is the overall risk trending up or down compared to past weeks?

The score gives you a starting point. The category breakdown tells you what's actually driving it. The live data layer (current advisory, weather, disaster alerts, recent news) tells you what's happening right now.

Why a composite score is dangerous if misused

A travel risk score is an estimate, not a prediction. It should never replace your own judgment, the current advisory from your country's foreign ministry, professional security advice for high-risk regions, or medical guidance from a travel-medicine specialist. It's especially limited in five common situations.

1. Rapid-onset events

Baseline scores update on a 90-day review cycle. Government advisories refresh every 6 hours. News headlines come in every hour. Even our fastest refresh layer cannot match the speed of a breaking event — a coup, an earthquake, a terror attack, or a sudden border closure. If something is unfolding live in the news, our score may not reflect it for hours. Always check the latest news for your destination before traveling.

2. Neighborhood-level variation

A city-level score smooths over enormous within-city variation. Mexico City Polanco is not Mexico City Iztapalapa. London Mayfair is not London Tower Hamlets at 2am. Use the score for "should I go" decisions; use neighborhood-level research and local advice for "where should I stay" decisions.

3. Activity-specific risks

A score for a city doesn't tell you about specific activities. Trekking, diving, motorcycle touring, climbing, and adventure sports carry risks that are nearly independent of the destination's city-level safety. A low-risk city can host a high-risk activity. Always research activity-specific safety separately.

4. Population-specific risks

Different travelers face different risk profiles. LGBTQ+ travelers face higher legal and social risks in roughly 70 countries where same-sex relationships are criminalized. Solo female travelers face higher street-harassment risks even in "safe" cities. Journalists, researchers, and aid workers face higher targeting risks in conflict zones. Our personal modifier captures some of this but cannot replace tailored advice.

5. Catastrophic tail risks

Composite scores collapse the distribution of outcomes into an expected value. A city with a 0.1% chance of a major earthquake and 99.9% chance of a normal trip will score similarly to a city with steady moderate risks across all twelve categories. The two trips are not actually equivalent. For tail risks, look at category scores individually — especially natural disaster, political, and health.

How our score compares to other travel risk indices

Several other organizations publish travel risk indices — International SOS, Riskline, Crisis24, and various travel insurance providers. Each uses a different methodology and weighting, which is why scores for the same city can vary widely. Here's how the major approaches differ.

Government travel advisories (free, conservative)

The US State Department, UK FCDO, Canada GAC, Australia Smartraveller, and similar government services publish travel advisories with 4–5 tier scales. These are authoritative but slow to update, often politically influenced, and rarely give category-level detail. They're the right answer to "should I avoid this country?" but a blunt instrument for "what specifically should I prepare for?"

Travel insurance risk ratings (free for policyholders)

Insurance providers (Allianz, World Nomads, Travelex) publish risk ratings tied to their coverage. These are useful for premium estimation but heavily weighted toward what they pay out on — usually medical and theft — with less emphasis on political or disaster risk.

Corporate security indices (paid, detailed)

International SOS, Crisis24, and Control Risks publish corporate-traveler indices with detailed category breakdowns and 24/7 monitoring. These are excellent but cost thousands per year and target enterprise clients.

Crowdsourced safety apps (free, fast, noisy)

Apps like Numbeo, GeoSure, and various Reddit travel communities crowdsource current conditions. They're fast but suffer from selection bias (who reports?), reporting bias (what gets reported?), and limited verification.

Our score (free, transparent, slower than commercial)

We're closest to the corporate-security approach in methodology but free, open about our weights, and built on public data sources. We're slower than crowdsourced apps and less comprehensive than paid services. The trade-off we make: transparency about how the number is built, every claim cited to a source, no paywall.

How to read a travel risk score in practice

The most common mistake is to look only at the composite number. Here's what to actually look at, in order.

1. The tier. Low / Moderate / Elevated / Extreme tells you the overall planning posture — standard precautions, informed planning, meaningful precautions, or reconsider travel.
2. The top three drivers. Every assessment lists the three categories pushing the score highest. These are what you should plan around. If "crime" is the top driver, learn the safe-neighborhood map and use registered taxis. If "health" is the top driver, see a travel-medicine specialist before departure.
3. The data confidence. A low-confidence score is a guess. A high-confidence score is built on city-specific baselines that have been reviewed in the last 90 days. Don't make decisions on low-confidence scores without backing them up with other research.
4. The current advisory. If your country's foreign ministry says "do not travel," that overrides any score from us or anyone else.
5. Recent headlines. Spend 90 seconds scanning the news box on the assessment page. If something is unfolding, you'll see it.
6. What is unavailable. Every assessment lists what we don't know about your destination. If something important is on that list, treat the score with extra caution.

Frequently asked questions

Is a travel risk score the same as a travel advisory?

No. A government travel advisory is an official recommendation from your country's foreign ministry to its citizens. A travel risk score is an analytical estimate that can be produced by any organization with appropriate methodology. Our score uses the advisory tier as one input (weighted 0.12) but is not a substitute for it. If your government says "do not travel," follow that.

Can a low score still be dangerous for me specifically?

Yes. Composite scores reflect average traveler experience. If you have specific risk factors — LGBTQ+ status in a country where that's criminalized, a chronic medical condition, traveling solo as a young woman in a country with elevated street harassment, journalism in a country with press restrictions — the score for you could be materially higher than the published score. The personal modifier captures some of this, but tailored advice from someone with knowledge of your situation is always better.

How often does a city's score change?

Baseline scores are reviewed every 90 days. Live data layers refresh every 30 minutes to 12 hours depending on the source. AI-synthesized briefings refresh every 6 hours per assessment. So in practice, the same city can show a noticeably different score from one week to the next if an advisory changed, a disaster occurred, or news broke. Re-run the assessment the day before you travel.

Can I trust an AI-generated travel briefing?

Within strict limits. Our AI prose synthesis is bound to specific cited sources — every claim links back to the snippet it was built from, and unsupported claims are dropped before publication. The AI does not set the threat scores; humans do. The AI does not invent the data sources; we feed it a fixed set of search results. See our AI Content Policy for the full set of guardrails.

Where does the score come from technically?

Curated city baselines plus public-data feeds (Open-Meteo for weather and air quality, the US State Department for advisories, GDACS for disaster alerts, the CDC for health notices, Google News for recent headlines) plus an AI synthesis layer for prose. See Methodology and Data Sources for the full pipeline.