Top 100 Countries by Crime Index in 2026

What the Numbeo Crime Index measures

Numbeo builds its Crime Index from survey responses submitted by website visitors. Questions cover perceptions of overall crime, whether crime is increasing, daytime and nighttime safety, worries about robbery, car theft and assault, property crime, violent crime, drugs, harassment and corruption or bribery.

Survey results are scaled to a 0–100 index. Numbeo describes scores above 80 as very high, 60.01–80 as high, 40.01–60 as moderate, 20.01–40 as low and 20 or below as very low.

The 2026 Mid-Year label identifies a semiannual historical snapshot. It does not mean every underlying survey response was submitted during 2026. Numbeo states that its current index uses recent survey data and that semiannual historical indices incorporate the latest available responses into periodic snapshots.

Because this is perception data, differences in participation, respondent mix, local expectations, reporting culture and sample size can affect comparisons. The country ranking page does not display a contributor count for every row, so identical decimal precision should not be interpreted as identical sampling precision across all entries.

Top 10 highest Crime Index scores

The ten highest scores range from 80.8 to 67.4. Papua New Guinea and Venezuela are the only entries in this Top 100 above 80, while Haiti is close behind at 79.8.

Top 10 entries by Numbeo Crime Index, 2026 Mid-Year

Ranking uses the published Crime Index value in descending order. When displayed values are equal, the source-table order is retained rather than applying an undocumented tie-break rule.

Methodology

The ranking uses Numbeo's published Crime Index for the fixed 2026 Mid-Year historical snapshot. The source value is measured on a 0–100 scale and sorted from highest to lowest. StatRanker does not recalculate the index, combine it with police statistics or create a projected 2026 crime value.

Numbeo states that its crime data are derived from surveys answered by website visitors. Individual responses are converted to numerical values, suspicious submissions are filtered and results are scaled from 0 to 100. The semiannual historical index is updated twice a year as new data are incorporated.

The 2026 Mid-Year label therefore refers to the published snapshot, not to a standardized calendar-year crime count. Numbeo's current index uses recent survey responses over a multi-year window, while historical semiannual snapshots preserve specific published comparison points.

The country ranking does not display a contributor count beside every country or territory. That limits assessment of sampling precision at row level and is one reason the index should be interpreted as a perception indicator rather than a substitute for official crime-rate statistics.

Official homicide statistics answer a different question. An intentional homicide rate, for example, normally expresses homicide victims relative to population, typically per 100,000 people. That type of administrative or public-health measure should not be merged with Numbeo's perception score.

The metric does not directly measure homicide, police-recorded offenses, victimization prevalence, convictions, conflict exposure, organized crime or an individual's probability of becoming a crime victim.

Main ranking: Top 100 Crime Index entries

Search by entity, filter by broad region or data status, change sort order, or limit the view to the Top 10 or Top 20.

Top 100 entries by Numbeo Crime Index, 2026 Mid-Year

Ranking values come from the fixed Numbeo 2026 Mid-Year historical snapshot. Regional labels are supporting classifications and do not alter the source value or rank. Thailand is the 100th entry on this page; the Numbeo source table continues below it.

What stands out in the Top 100

How to interpret this ranking

This table is useful for comparing perceived crime and safety conditions across countries and territories under one consistent Numbeo methodology. A higher score means respondents produced a higher Crime Index; it does not establish that the same place has the world's highest official crime rate.

For harder comparisons of lethal violence, an intentional homicide rate per 100,000 people is a different and narrower metric based on victims relative to population. StatRanker's separate homicide rate by country ranking should therefore be read independently rather than merged with this perception index.

Readers looking specifically for low perceived urban crime should also distinguish country-level data from city-level data. StatRanker's 2026 safest cities comparison uses city-level safety data and answers a different geographic question.

The main interpretation risk is treating a broad perception index as if it were an administrative crime count. Both can be useful, but they measure different things and can move differently.

FAQ

Sources