Wandroz

← London

Westminster, London

🟢 Official crime data · Police-recorded crime, data.police.uk · May–July 2026 Methodology →

Relatively safer day Relatively safer night

Relative rank 5 among boroughs currently covered. Based on official Metropolitan Police data, averaged over 3 months (2026-05 to 2026-07).

Real administrative boundary — Office for National Statistics 2021 boundaries, not an approximate circle.

Recorded incidents (total across 3 months, 2026-05–2026-07)
CategoryCount
Anti social behaviour3615
Other theft3585
Violent crime3432
Theft from the person3022
Shoplifting2629
Public order1049
Vehicle crime651
Robbery605
Burglary598
Drugs578
Bicycle theft314
Possession of weapons130
Other crime125
Criminal damage arson41

20374 incidents total over 3 months — averages 6791.3/month, which is what the day/night scores below are based on.

Data completeness: full months, real administrative boundary — 20374 recorded incidents total across 3 months (2026-05–2026-07), queried against Westminster's actual ONS 2021 polygon outline (the same shape shown on the map above), not an approximate circle. This is the complete monthly response from the UK Police API for each month in that window. Day score: 1.6031 weighted incidents/1,000 people, averaged over the last 3 months (63% of the 33-area average). Night score: 9.4626 weighted incidents/1,000 people, same window (75% of the 33-area average). Both are normalised by an estimated workday/footfall population of 747,552 rather than the 2021 resident population of 204,300, so central boroughs with heavy commuter and visitor footfall aren't overstated as riskier simply for having fewer official residents. The footfall estimate is derived from the ONS's 2011 Census workday-population release — the most recent one published — applied to Westminster's current resident population, and it is therefore an approximation of today's footfall, not a measurement of it.

What that rating does not say. Westminster records more thefts from the person than almost anywhere else in London — 3,022 in the three months behind this rating. The green above is a rate: fewer weighted incidents per person estimated to be present than in most boroughs, because so many people are present. It is not a statement that little is stolen here, and it does not mean a low chance of being pickpocketed on a crowded street. Treat crowds here the way you would anywhere with this much footfall. Rated against the average of the 33 areas currently on the automated pipeline — 32 boroughs and the City of London.
Search accommodation here on Booking.com →

This link is already scoped to Westminster, not the whole of London.

Source: UK Police street-level crime data →

Frequently asked questions

Is Westminster safe?

Based on real, current UK Police data, Wandroz rates Westminster as relatively safer than most other neighbourhoods in London during the day and relatively safer than most other neighbourhoods in London at night, relative to the other London boroughs currently on the automated pipeline. The rating comes from 20374 recorded incidents over 3 months (2026-05 to 2026-07), split into a day-weighted score (property crime) and a night-weighted score (violence, robbery, antisocial behaviour).

Is Westminster safe at night?

Westminster's night score is relatively safer than most other neighbourhoods in London, based on the categories most relevant after dark (violence, robbery, street theft, public order, antisocial behaviour), normalised by an estimated workday/footfall population rather than plain residents where that data exists, so busy central boroughs aren't overstated as riskier just for having fewer official residents.

What official data is Westminster's rating based on?

20374 recorded incidents from data.police.uk (the UK Police's official open crime API), queried against Westminster's real administrative boundary and covering 3 months (2026-05 to 2026-07) — recomputed from the source data whenever the Met publishes new months, not a one-off snapshot. Full category breakdown is shown further down this page.

See something wrong or out of date on this page? Report a correction →