This page explains how we turn official Statistics Canada figures into the guides and rankings on Canada Crime Map: which measures we use, how we rank places, what arithmetic we do ourselves, and where the numbers have limits. Our sources and licences are listed on the data sources page.
Current data: Statistics Canada, police-reported crime statistics for 2025, released 22 July 2026. Updated 5 October 2026.
The two main measures: Crime Severity Index and crime rate
We usually show two numbers for each place. They answer different questions, so it helps to read them together.
Crime rate
The crime rate is the number of Criminal Code incidents (excluding traffic offences) reported to police per 100,000 people. Every incident counts the same: a shoplifting and a homicide each add one. In 2025 Canada’s crime rate was 5,585 per 100,000.
The crime rate tells you how much reported crime there is. Because minor, high-volume offences such as theft and mischief make up most incidents, it is driven mainly by those.
Crime Severity Index (CSI)
The Crime Severity Index also measures reported crime per person, but it gives each offence a weight based on how serious it is, using the average sentences courts hand down. A homicide or robbery therefore counts for far more than a minor theft. The index is set so that Canada in 2006 equals 100. In 2025 Canada’s CSI was 75.0.
Statistics Canada also publishes a violent CSI and a non-violent CSI. We use these where they help explain what is driving a change.
We rank places by CSI because it reflects both how much crime there is and how serious it is. But Statistics Canada itself warns that the CSI is not a universal indicator of safety. A place can have a high CSI because of a small number of very serious incidents, or because of a large volume of property crime concentrated in a commercial area. Our longer explainer, Crime Severity Index explained, goes into more detail.
Metro areas vs police services
Most of our city figures are for a census metropolitan area (CMA): a whole metro region made up of a core city and the surrounding municipalities linked to it by commuting. The Toronto CMA, for example, includes many suburbs outside the city of Toronto. Statistics Canada published figures for 41 CMAs in 2025.
For some places we also use figures for a police service, which covers the area that service polices. This can be smaller than the CMA (a city police force) or cut across it differently (a regional police service such as Peel or York). The two can give quite different results:
| Area (2025) | Type | Crime Severity Index |
|---|---|---|
| Toronto | Census metropolitan area | 55.0 |
| Toronto Police Service | Police service (city of Toronto) | 60.0 |
| Victoria | Census metropolitan area | 74.6 |
| Victoria Police Department | Police service (Victoria and Esquimalt) | 161.2 |
Source: Statistics Canada, Tables 35-10-0026-01, 35-10-0188-01 and 35-10-0063-01, 22 July 2026.
Downtown cores usually record more crime per resident than their suburbs, partly because many people who work, shop, travel or go out there do not live there, while the rate is calculated per resident. We always say which geography a figure refers to. Where policing changed during a year (for example, Surrey’s transition from the RCMP to the Surrey Police Service), we explain the split rather than add figures that are not comparable.
How our rankings are made
- We take the official CSI for every area in the comparison (for example, all 41 CMAs, or all CMAs in one province) for the same year.
- We sort them by CSI, from highest to lowest for “most dangerous” lists and from lowest to highest for “safest” lists.
- Ties: where two places show the same CSI to one decimal place, we order them by the unrounded value Statistics Canada publishes. For example, the Toronto CMA (54.99) and the Québec CMA (54.97) both round to 55.0 for 2025, so Toronto is placed just above Québec. If the unrounded values are also equal, we list them in alphabetical order and say they are tied.
- We show the crime rate and the year-on-year change alongside the CSI so that readers can see whether a ranking is driven by volume, severity or a one-year jump.
- We do not mix geographies in one ranking. CMAs are ranked against CMAs, provinces against provinces and police services against police services.
Words such as “most dangerous” and “safest” in our titles describe position in these official rankings. They are relative to other Canadian places, not a statement that any place is dangerous or safe in absolute terms.
Our own arithmetic
Statistics Canada publishes the CSI, crime rates, counts and year-on-year changes. Some comparisons in our articles are our own calculations from those published figures, and we say so on every page. They include:
- Difference from the national average: (place value − Canada value) ÷ Canada value × 100. For example, the Calgary CMA’s CSI of 61.7 against Canada’s 75.0 works out at about 18% below the national figure.
- Changes over several years: (later value − earlier value) ÷ earlier value × 100, using the published series.
- Ratios: how many times higher one rate is than another.
We use the unrounded published values for these calculations and round only the result. Pages that include our own calculations use the “Adapted from Statistics Canada” attribution.
Rounding
- CSI values are shown to one decimal place (for example, 75.0).
- Crime rates per 100,000 are shown as whole numbers (for example, 5,585); rates for rare offences such as homicide are shown to two decimal places (for example, 1.61).
- Percentage changes are rounded to the nearest whole per cent. Where Statistics Canada publishes a rounded percentage change, we use its figure.
- Because of rounding, a percentage we calculate may differ by about one point from one calculated from the rounded figures shown.
Small populations and volatility
In a smaller metro area, a few serious incidents can move the CSI sharply from one year to the next, and homicide counts in single digits can double or halve by chance. We point this out where it matters and, where possible, show several years of data so a single unusual year is seen in context. A one-year change in a small place should be treated with caution.
Other limits
- Reported crime only. Many crimes are never reported to police; in 2019 Statistics Canada’s victimization survey estimated only about 29% were. Changes in reporting or in police practice can affect the figures.
- Annual data lag. Each year’s figures are published the following July.
- No neighbourhood statistics. Statistics Canada does not publish neighbourhood-level figures. We do not estimate them and we do not label neighbourhoods as dangerous.
- Places, not people. Police data describes incidents in an area. It says nothing about the people who live there.
What we plan to add
Interactive maps
We are building maps from police open data published under open licences (see data sources). Incident locations in this data are offset by police to the nearest block or intersection, so the maps will show patterns, not exact addresses. Each map will show its data source, licence and the date it was last updated.
Safety Score (future)
We may introduce a “Safety Score” to summarise official data for an area or route. If we do:
- it will be our own estimate, adapted from official data, and not an official statistic;
- we will publish its full method on this page before it goes live, including which data it uses, how offences are weighted and how small areas are handled;
- it will be labelled clearly wherever it appears.
Update schedule
- Annual statistics: we update city guides, rankings and national pages after Statistics Canada’s annual release each July. The next release, covering 2026, is expected in July 2027.
- Revisions: if Statistics Canada revises earlier years, we update the affected pages.
- Maps: once live, maps will be refreshed when the police services update their open data.
- Every article shows the date it was last updated.
Questions about our method? Email ggshop@safety-app.uk. See also our editorial policy and disclaimer.