Crime in London: what the police data actually shows

London is policed by one force covering 32 boroughs plus the City of London, and it produces more headlines about crime than anywhere else in the country. Most of those headlines boil a borough down to a single ranking. The published numbers are more interesting than that, and more useful, if you read past the ranking.

The boroughs with the highest and lowest recorded rates

Using Metropolitan Police recorded-crime figures for 2024, the boroughs with the highest crime rate per 1,000 residents were Westminster (349.9 per 1,000, around 91,000 offences), Camden (172.4 per 1,000, around 42,800 offences), Kensington and Chelsea (145.7 per 1,000), Islington (127.6 per 1,000) and Southwark (122.7 per 1,000). The lowest were Richmond upon Thames (60.7 per 1,000), Kingston upon Thames (62.1 per 1,000) and Sutton (62.9 per 1,000).

Why that ranking is misleading on its own

Westminster’s rate looks alarming next to Richmond’s until you notice what “per resident” means. Westminster’s registered population is small relative to how many people are actually there on an average day — office workers, shoppers, tourists, commuters passing through the West End and the entertainment district. A crime rate calculated against resident population massively overstates the risk to any one resident, because the denominator excludes almost everyone who is physically present. Central, tourist-heavy and commercial boroughs will structurally always rank “worst” on this measure, regardless of how safe they actually are to live in. Theft was the most commonly reported offence in these central boroughs, which fits: pickpocketing and phone theft cluster where footfall is highest, not where residents are most at risk.

The outer boroughs tell a different story. In areas like Sutton, violence against the person made up a larger share of recorded crime than theft did — a genuinely different pattern, worth knowing if you’re weighing up personal safety rather than just a headline number.

What this means if you’re actually trying to judge an area

A single borough-wide rate flattens huge internal variation. A borough’s overall number tells you almost nothing about a specific street, and the same is true in reverse — a “safe” borough can have a genuinely higher-risk pocket a road-length or two behind a main tourist strip. If you’re checking an area you might live in, work in, or send a child to, a borough league table is the wrong tool. Our address checker pulls twelve months of official police data for the actual postcode you care about, not the borough it happens to sit in, and breaks it down by category rather than a single combined number.

Compare well-known London areas yourself

A snapshot number goes stale the moment it’s published. The more useful move is to pull real, current twelve-month data for two specific places at once and read the actual category mix — not just the totals. A few comparisons worth trying, each opens live with both postcodes already loaded:

• Westminster vs Camden Town — a government-and-tourist district against a nightlife-and-market district. Even if the totals end up close, expect the category mix to look very different: Westminster’s footfall is daytime and transient, Camden’s leans evening and weekend.
• Notting Hill vs Canary Wharf — an affluent, largely residential and market area against a weekday financial district that empties out at night and at weekends. A place that’s only busy nine-to-five reads very differently across a full week than one that’s residential around the clock.
• Shoreditch vs Westminster — two of the capital’s most-visited areas, for almost opposite reasons: nightlife and creative industry versus government and mass tourism.

None of these comparisons tells you which area is “worse.” What they can show you is whether a place’s crime profile matches what actually happens there day to day — and whether the categories that matter most to you (burglary if you’re renting, violent crime if you’re out at night, vehicle crime if you’re parking a car) sit where you’d expect, or somewhere you wouldn’t. That’s a more honest way to judge an area than any single ranking.

A live snapshot from the police API (July 2026)

To make that concrete, here is what the categories actually looked like in a small area around each postcode last month, pulled live from data.police.uk — the same feed the address checker uses — and checked by hand against the raw record count. The sample areas are deliberately small, roughly 250–300 metres across for the four residential and nightlife postcodes, and a wider radius for Westminster: the police API caps a single request at 500 records and silently truncates the rest once a busy area hits that limit, so keeping each box small enough to stay under it, and checking the totals by hand, is what lets us show complete real figures instead of a truncated guess.

Read these as a mix, not a scoreboard. The percentages show what share of nearby incidents fell into each category — not which area had “more” crime overall, since the sample areas aren’t the same size on purpose (a raw side-by-side total would be exactly the kind of misleading comparison this article is arguing against). What’s worth noticing: Canary Wharf and Westminster both skew towards violent crime and theft from the person, consistent with commuter and night-time-economy areas, while Notting Hill and Camden Town lean more towards shoplifting, and Shoreditch is dominated by theft from the person — in keeping with a dense nightlife and tourist district. This is one month of data, not a trend. For the full twelve-month picture and an actual up-or-down trend for one address, use the checker itself.

The limits of this data

This is recorded crime — what was reported and logged, not everything that happened. It reflects a single year and a single measurement (crimes per resident), not the only way to measure risk. And a rate, high or low, is not a verdict on a place or the people who live there; it’s one data point among several worth weighing. If you want the fuller picture of how UK police data works and what it can’t tell you, we’ve written that up separately.

And as always: none of this replaces calling 999 in an emergency. This is context for everyday decisions, not a substitute for the emergency services.