If you have ever tried to look up UK crime figures, you have probably hit two numbers that disagree. The Office for National Statistics says one thing, the police.uk map shows another, and journalists quote both. Neither is wrong — they measure different things. Here is what each source actually does, and which one to trust for which question.
What the ONS Crime Survey Measures
The Crime Survey for England and Wales (CSEW), published by the ONS, is a household survey. It asks tens of thousands of people each year whether they have been a victim of crime in the past twelve months, regardless of whether they reported it to the police. The survey is the standard source for the long-run national crime trend, and it captures the so-called "dark figure" of unreported crime that recorded-crime data misses entirely.
What it does not do: tell you about your postcode. The survey samples are designed to be representative nationally and at broad regional level, not at street, postcode or even ward level. It also excludes some categories outright, such as fraud against businesses, and historically excluded under-16s.
What Police.uk Data Measures
Police.uk publishes recorded crime — every offence that police forces actually logged in a given month, with a category and an approximate location. It is the source most postcode-level tools, including ours, are built on. Our police.uk crime map guide covers how to read it directly.
What it does not do: capture unreported crime. If victims do not call the police, those incidents never enter recorded crime. It is also sensitive to changes in police recording practice — a force tightening its standards can lift recorded counts without any change in actual offending. Our explainer on what UK crime statistics mean goes into this in more depth.
Why the Two Sources Disagree
The gap between ONS and police.uk numbers is mostly about reporting. Bicycle theft, low-value vehicle damage and minor anti-social behaviour are often not reported, so the ONS estimate is higher than recorded crime. For homicide, where almost every case enters the official record, the two sources line up closely. For violence and fraud the gap is wider, and the direction of any change can even differ between the two for a year or two before converging.
Which One to Use for Which Question
- "Is crime rising in the UK?" ONS Crime Survey. The recorded-crime trend is too sensitive to recording-practice changes to answer this on its own. We cover the headline-trend question in our piece on whether UK crime is really rising.
- "How much crime is on my street?" Police.uk. The ONS survey cannot answer postcode-level questions; only recorded crime is published at that resolution.
- "How does my area compare with the one next door?" Police.uk, with the caveats in our two-postcode comparison guide.
- "What share of crime gets reported?" ONS, by category. The survey is the only source that can tell you the reporting rate.
Treat Them as Complementary
The two sources are not competing — they are answering different questions. Use the ONS for the long-run national picture and to understand reporting rates. Use police.uk for anything postcode-level. A CrimeSafe report builds on the police.uk data, pulling 24 months of recorded crime for any postcode with full category and outcome detail, and a safety score for direct comparison. See our national safest-areas rankings for wider context.