Buyer Guides11 min read

Checking crime before you buy: what the Police.uk data really shows (and what it does not)

Police.uk publishes street-level crime data for England, Wales and Northern Ireland — but it is recorded crime, snapped to anonymised points, and useless to compare raw between areas. Here is how to read it honestly before you make an offer.


TL;DR


Crime is one of the first things buyers search for — and one of the easiest to misread. The open data behind almost every "is this area safe?" tool is Police.uk, which publishes street-level crime points for England, Wales and Northern Ireland (Police Scotland does not contribute, so Scottish coverage is absent).


Three things you must hold in your head before you trust a single number:


1.It is recorded crime, not actual crime. Reporting rates vary hugely by offence and by area.
2.The street-level points are snapped to anonymised locations — usually the centre of a street or a nearby landmark — not the exact address where the crime happened.
3.Raw counts are meaningless to compare between areas until you normalise them. A busy town centre will always "look" worse than a quiet cul-de-sac a mile away, because it has more people, more shops and more footfall — not necessarily more risk to a resident.

Used carefully, the data tells you about the *texture* of an area: the mix of offence types, whether the trend is rising or falling, and how a residential street compares to like-for-like streets.


Where the data comes from


Police.uk is the official open-data service run on behalf of the Home Office and the police forces of England, Wales and Northern Ireland. Every month, each force uploads the crimes recorded in its area, released roughly six to eight weeks in arrears. Each record carries an offence category, a month, an anonymised location (a human-readable street name like "On or near Acacia Avenue"), and, where available, an outcome.


Coverage note worth repeating: Police Scotland does not submit to Police.uk. If you are buying in Scotland, this dataset is silent, and you should not infer "low crime" from an empty map — it is an *absent* map.


The crime categories


Knowing the categories matters, because the *mix* tells you far more than the total. Ten burglaries say something very different about a residential street than ten counts of shoplifting.


CategoryWhat it coversWhy it matters to a homebuyer
BurglaryBreak-ins to homes and other buildingsThe single most relevant category for a resident
Vehicle crimeTheft of/from vehicles, interferenceRelevant if you park on-street
Anti-social behaviourNuisance, rowdy or inconsiderate behaviourVolume category; quality-of-life signal
Violence and sexual offencesA broad combined category, minor to seriousLarge bucket; concentration near nightlife
Criminal damage and arsonVandalism, graffiti, deliberate firesStreet-level wear-and-tear signal
Theft from the personPickpocketing, snatch theftFootfall-driven; clusters in busy areas
ShopliftingTheft from shopsTells you there are shops, not unsafe homes
RobberyTheft with force or threatLower volume; clusters near transport/retail
DrugsPossession and supply offencesPartly reflects police activity, not just prevalence
Public orderThreatening or disorderly behaviourOften nightlife-adjacent
Bicycle theftSelf-explanatoryHigher near stations and student areas

The combined "violence and sexual offences" category is the most often misread. It bundles a huge range of seriousness into one label, and a large share concentrates around pubs, clubs and transport hubs — which is why a flat above a high street can show a frightening-looking count that has little to do with the safety of the flat itself.


The four ways people read this data wrong


1. Treating recorded crime as actual crime


Police.uk shows crimes reported to and recorded by the police. The Crime Survey for England and Wales (run by the ONS) consistently finds that many offences go unreported. Reporting is also shaped by confidence in the police and insurance requirements. So a higher recorded count can reflect *better reporting*, not more crime.


2. Forgetting the points are snapped, not exact


To protect victims' privacy, Police.uk maps each incident to the nearest anonymised "snap point" — typically the centre of a street or a nearby landmark. The dots do not mean "a crime happened at this exact spot". Zooming to a single house and reading the dots literally is a mistake.


3. Comparing raw counts between areas


Raw monthly counts are driven first by how many people and how much activity an area contains. Before any comparison is meaningful you have to normalise: ideally crimes per head of population (using a denominator like the ONS Census 2021 count), and always like-for-like — residential against residential, not a quiet road against a town-centre postcode. This is the single most common error in DIY crime research.


4. The one-mile-radius trap


Many quick lookups draw a circle — often a one-mile radius of a town centre — and tally everything inside it. If that circle catches a high street or a station, the total is dominated by shoplifting, theft from the person and footfall-driven public-order offences that have nothing to do with a residential street. A home 800 metres from a busy centre can inherit a terrifying number that belongs to the shops. This is exactly why we frame crime the way we do on a postcode area page: the headline is the *category mix* and how the area compares to like-for-like areas, not a raw radius count.


How to actually read it before you buy


+Start with the mix, not the total. Is the crime dominated by shoplifting and public order (a retail signature) or by burglary and vehicle crime (a residential signature)?
+Look at burglary and vehicle crime specifically — the categories most tied to residential risk and most consistently reported.
+Normalise before you compare. Use a per-capita rate or a tool that does it for you.
+Read the trend, not one month. Look across at least a year.
+Check the outcomes. A high share of "investigation complete; no suspect identified" is normal for volume crime.
+Ignore the exact dots. Read the map as a rough heatmap, never as a record of specific addresses.
+Triangulate. Cross-reference against the Crime domain of the Indices of Multiple Deprivation 2025, and your own visit at different times of day.

Crime is one layer of due diligence


A sensible pre-offer check looks at crime *alongside* the other official datasets: flood risk (Environment Agency), planning (planning.data.gov.uk), schools (Ofsted / GIAS), broadband (Ofcom), road safety (DfT STATS19), environmental factors (UKHSA radon, DEFRA air quality, Historic England), and sold-price history (HM Land Registry Price Paid, England and Wales only). Pulling those together for a single address is what a desk due-diligence report does — ours is a £14.95 one-off with no subscription. Run any postcode from search, see the area context on an area page, or check what's included on the pricing page.


Honest caveats


+This is recorded crime in England, Wales and Northern Ireland only. Scotland is not covered. Absence of data is not evidence of low crime.
+Locations are anonymised and snapped. The map shows approximate clusters, not exact addresses.
+Counts must be normalised. Raw totals reflect population and footfall as much as risk.
+Reporting rates vary, so recorded crime understates and unevenly samples actual crime.
+There is a reporting lag of roughly six to eight weeks.
+No dataset predicts the future. Crime data describes what was recorded, not what will happen on your street after you move in.

Used with these limits in mind, Police.uk is genuinely useful — it just rewards the buyer who reads it like an analyst, not like a headline.


Ready to research a property?

A full property dossier for £14.95. Paid once, no subscription.

Search for a property →