INCIDENTMTL

Method, sources and licences

Every figure on this site is a count of published data. This page explains where it comes from, which periods it covers and how we process it.

These are not official statistics of the SPVM or the City of Montréal.

Data shown now

  • Area pages cover the twelve months from October 2025 to September 2026.
  • SPVM reports published up to 11 October 2026.
  • Monthly sector totals up to September 2026.
  • SIM interventions up to September 2026.
  • 311 requests up to September 2026.
  • Press coverage read up to September 2026.

Sources

The City of Montréal’s open data is published under the Creative Commons Attribution 4.0 licence (CC BY 4.0).

City of Montréal open data used under CC BY 4.0, modified and interpreted by IncidentMTL. The City of Montréal does not endorse this site.

Periods

Pages for neighbourhoods, boroughs, linked towns and police sectors count the last twelve complete months; a month counts as complete seven days after it ends. Each figure is compared with the twelve months before.

Yearly totals start where each source’s data starts: 2015 for SPVM reports and press coverage, the first month of each request type for 311 requests (2022 at the earliest), and the first month we hold for SIM interventions. A year the data covers only in part, such as the current year, is marked as partial.

Address pages show the last 7, 30 or 90 days (90 by default), within 500 m, in Montréal days, from the same copy of the data the IncidentMTL app reads. SIM interventions are not shown there.

Locations

The SPVM moves each report to an intersection on the street segment where the event happened, to protect privacy. A point on our maps is therefore a nearby intersection, never a building.

SIM interventions and 311 requests are placed as the City publishes them. Press-reported events are placed as the articles describe them.

PDQ 50, the SPVM’s métro unit, is published with fictitious positions; it is left out of every figure, map and ranking.

The SPVM publishes only six categories as points: break-ins, vehicle theft, theft from vehicles, mischief, robbery and offences causing death. Assaults, sexual assaults and fraud, among others, appear only in the sector totals.

Areas and residents

Neighbourhoods are the City’s “Quartiers de référence en habitation”; each belongs to one borough or linked town. A borough or linked town made of a single neighbourhood has a single page.

Residents and dwellings come from the 2021 Census dissemination blocks. Each block counts whole in the neighbourhood that contains its centre or, at the edges, the nearest one; a borough or town adds up its neighbourhoods, and the island its boroughs and towns. For police sectors, a block cut by a boundary is shared by the part of its area on each side.

Rates are per 1,000 residents over twelve months. They divide by the people who live in the area: where many people work or pass through without living there, such as downtown, the rate per resident reads high.

Rankings

A rate per resident and a rank need at least 2,000 residents. A smaller area shows its counts, with no rate or rank, and says why.

The 34 boroughs and linked towns are ranked among themselves, neighbourhoods among all neighbourhoods with their own page, and police sectors among the 28 sectors.

Rank 1 goes to the highest rate per resident for the six SPVM categories together, over the last twelve months. A rank compares areas; it says nothing about any street or address.

Police sector totals

Police sector pages lead with the monthly totals the SPVM publishes for each sector, in more categories than the points: the twelve months up to the latest month published, compared with the twelve before, by group and by category. These are sector totals, not incident locations.

Press

Pages count serious incidents reported in the press: shootings, stabbings, beatings and killings (the categories assault, shots fired or a firearm, and offence causing death). Deaths reported with no cause stated and deaths during a police intervention are not counted.

Articles are grouped into events. An event the articles place to at least the street (a street, a block or an intersection) is pinned there, and counts in the neighbourhood, the borough or town and the police sector that hold its pin; where the pin and the borough the articles name differ, the pin wins. An event without a pin counts only for the borough or linked town the articles name. A pinned event whose pin falls in no borough also counts for the borough or town the articles name. Each event counts once in each area, and the island counts them all.

Coverage goes back to January 2015 and runs to the last month our archive has read for every outlet it follows; each page states that month. Headlines are shown as published, with the outlet and the date, and link to the article.

Maps

Area maps group the last twelve months’ SPVM reports into hexagons of 250 m for neighbourhoods and sectors, 500 m for boroughs and towns, and 1,000 m for the island.

Limits

A report is not always a confirmed crime, and many crimes are never reported: according to Statistics Canada (2019), about half of break-ins, vehicle thefts and robberies are not.

These figures are not official statistics. If you see an error, tell us: corrections are free. Report an error

This is historical public data. In an emergency, call 911.

Data sources

City of Montréal open data, used under CC BY 4.0, modified and interpreted by IncidentMTL. The City of Montréal does not endorse this site.

Population and dwellings: Statistics Canada, 2021 Census. Adapted from Statistics Canada, 2021 Census of Population: dissemination block boundary file and Geographic Attribute File (92-⁠151-⁠X), 2021. This does not constitute an endorsement by Statistics Canada of this product.

Press articles belong to their outlets; we cite the headline, the outlet and the date, with a link.