See an area before you visit it

How to use it

1. Pick a spot. Search a place, postcode or address (in the UK also a street), or tap the map. Outside the UK the circle goes up to 100 km.
2. Set the radius. Choose how far around the pin to look, and what the map colours show.
3. Read the scores. 10 is best, 1 is worst. The number, the word and the colour always go together, and the ? beside a score says what it measures.

1. 10Excellent
2. 9
3. 8Good
4. 7
5. 6Fair
6. 5
7. 4Weak
8. 3
9. 2Poor
10. 1

A guide, not advice

Crime Rates Map is a guide, not advice. It describes recorded crime in the past and how near services are. It cannot predict what will happen to you, and it is no substitute for visiting an area, talking to neighbours or asking a professional.

The Safety score

The Safety score is about crime only, and about how serious the crime is, not how many records there are. Police forces publish the location of each crime they record; we place every crime in the neighbourhood it happened in (an LSOA in England and Wales, a Data Zone in Northern Ireland). Each crime is then weighted by the Office for National Statistics Crime Severity Score: the typical sentence, in days, given for that kind of offence. A robbery counts 994, a burglary 465, violence and sexual offences 296, drugs 214, possession of weapons 177, vehicle crime 71, criminal damage and arson 25, public order 19 and shoplifting 13, so one robbery weighs as much as about 76 shoplifting offences. These are the ONS weights per police.uk crime category (edition: year ending March 2025; the mix of offences inside each category is averaged over all recorded crime in England and Wales in 2025/26).

Crime comes at two levels of detail. The map, neighbourhoods, postcode districts and the Safety score use the 14 broad categories police.uk gives each street-level crime, such as vehicle crime or criminal damage and arson. Council pages add detailed types (cars stolen, cars broken into, vandalism to homes, homicide, carrying a knife and others) from the Home Office's police recorded crime tables, which count every crime the police recorded in each community safety partnership area over a financial year. Those come only by council area, never by street: no source publishes murders or stabbings at street level. Knife crime is published only for whole police force areas.

Postcode district, town and council pages also show what happened to crimes reported 12 to 24 months earlier. The charge figure counts crimes where police.uk records that someone was charged or summonsed, including cases later found not guilty. We use crimes at least a year old because a recent crime has not had time to reach an outcome; even so, each figure is the latest outcome recorded so far, so it can still rise. Neighbourhoods have no outcome figure because their numbers are too small to be reliable.

Add up the weights of every crime in a neighbourhood and divide by its residents and the years covered: that is its severity per resident per year. We show it as a multiple of the average for the local authority the neighbourhood is in, worked out the same way over all its neighbourhoods, followed by the England and Wales figure ("1.3× the Chesterfield average (2.2× the England and Wales average)"): local first, because that is the comparison that matters when choosing between nearby streets. For context we also show the plain number of recorded crimes per 1,000 residents, with the types that make up most of the severity.

Anti-social behaviour is not a crime, and the ONS leaves it out of its own score, so it is not part of the Safety score. It is shown on its own line as incidents per 1,000 residents a year.

The severity per resident is ranked against every neighbourhood (or, for a postcode district, every district). The safest tenth scores 10, the next tenth 9, and so on down to 1 for the tenth with the most serious crime.

Police.uk moves every crime to the nearest of about 750,000 anonymised map points, such as the middle of a street, a pub or a shop, so a street on a boundary would put all its crimes in one neighbourhood and none in the one across the road. Crimes recorded on a map point within 30 m of a neighbourhood boundary are therefore shared equally between the neighbourhoods on each side, because the point usually sits on the boundary road itself. A neighbourhood's Safety score also blends its own severity per resident (three quarters) with the average of the neighbourhoods whose boundaries touch it (one quarter, weighted by residents). Busy centres are left out of the blend. The blended figure is ranked against every neighbourhood in the same way, and wherever the neighbourhood's own score differs, the map and the pin panel show it too. Postcode district, town, council and map circle scores are not blended: they already cover many neighbourhoods. The crime figures themselves (rates, types and trends) are not blended, although they include the neighbourhood's share of crimes on its boundary roads.

Every score on the site runs from 1 to 10, and 10 is always the best. Words go with the numbers: 9-10 excellent (for safety, very low crime), 7-8 good, 5-6 fair (average crime), 3-4 weak (above-average crime) and 1-2 poor (high crime, the worst fifth of neighbourhoods). The colour runs from green to red, and the number is always shown too.

How accurate are the scores?

We scored 37,503 neighbourhoods on police data from September 2016 to August 2023, then compared each with the score its own crime earned from September 2023 to August 2026. Town and city centres, and areas without enough police data, were left out.

79% were within 1 point. 39% scored the same. 6% were 3 or more points out.

Crime changes over time, so a perfect match is not expected. Both periods share crimes on boundary roads in the same way.

Checked against official figures: for 98% of the 252 councils we can compare, our total for April 2025 to March 2026 is within 10% of the Home Office's recorded crime for the same 12 months (neither total includes fraud, and ours leaves out anti-social behaviour, which is not recorded as a crime).

Why our scores can differ from other sites

• We score the area around you, not a single point on the map. Police.uk moves each crime to the nearest of about 750,000 points, so one postcode often has none and looks crime-free.
• We count crime per resident, so places of different size can be compared.
• We use 10 years of police data, weighted by how serious each crime is (the ONS Crime Severity Score).
• We update to the latest month of police data.
• Anti-social behaviour is shown but not scored, because it is not recorded as a crime.

Supermarket and transport scores

The supermarket score looks at the distance from the middle of each neighbourhood to the nearest large supermarket (over about 1,400 square metres) and to the nearest shop of any size. The transport score counts bus stops within 400 metres and adds the distance to the nearest railway station and tram or metro stop. Distances are in a straight line, not along the street, and a stop on the map is not a promise of a service. Both scores use fixed distance bands, from 10 (excellent access) to 1 (poor access), rather than a ranking against other places; a district score is the average of its neighbourhoods, weighted by population.

School and health scores

The schools score takes the best Ofsted outcome among primary schools and among secondary schools within 2 km of the middle of each neighbourhood, and averages the two. If there is no rated school of a kind within 2 km, that half counts as 1, and the page says so. All-through schools count as both. Ofsted changed its inspections in November 2025: schools with a new report card are shown by the middle of their three graded areas (and a failed safeguarding standard caps it at the lowest outcome), schools not yet inspected that way keep their last overall grade, a later ungraded inspection that confirms a grade (for example "School remains Good") counts as that grade, and the newest of these is used. A school that has not been inspected since it changed (for example, became an academy) is shown with its predecessor's outcome, and the page says so. A district's schools score is the average of its neighbourhoods' scores, weighted by population, and it lists the best rated schools within 2 km of any of its main neighbourhoods. Schools Ofsted has not rated are left out. Ofsted only inspects schools in England, so Wales, Northern Ireland and Scotland have no schools score.

The health score looks at the distance to the nearest GP surgery (the closer, the higher the score) and to the nearest hospital with an A&E department (likewise, over much longer distances), and averages the two. It measures how near care is, not how good it is or whether a practice is taking new patients. Care Quality Commission ratings of GP surgeries are shown for England only. Hospitals mapped in OpenStreetMap only as complex areas may be missing.

The Living score

The Living score is a weighted average of four theme scores: schools 33%, transport 25%, health 25% and supermarkets 17%. Crime is not part of it: it is the separate Safety score, shown beside it, and the two are never blended. When a theme has no score (schools in Wales and Northern Ireland, for example) it is left out and the weights of the rest are scaled up, and the page says so. The result is rounded to one decimal; the colour and the word beside it follow the nearest whole score. The weights are a judgement, not a measurement: a family may care far more about schools than about shops, so read the themes as well.

Some police forces stopped publishing data years ago (Greater Manchester Police after June 2019, for instance). Where a district's crime data ends more than a year before the newest month, the page says so; the Safety score then describes the years before that.

The map pin

On the home page you can drop a pin anywhere in the UK and choose a radius from 100 m to 10 km. The neighbourhood under the pin is shown first, with its scores and its own population. The circle around the pin is a separate section: it covers every postcode within the radius, which usually means parts of several neighbourhoods. Each neighbourhood's people, and its crimes, are shared out over its live postcodes, and the circle counts the share of each neighbourhood's postcodes that falls inside it. People live where postcodes are, so fields, parks and industrial land add nobody. The crime figure is then per resident like everywhere else on the site. Schools, shops, stations, stops and GP surgeries are those whose straight-line distance from the pin is within the radius; the nearest hospital with A&E is shown even when it is further away. Where a police force published fewer months than others in the circle, the panel says so. In Scotland there is no Safety score yet; Living covers it, so a Scottish pin shows Living and its themes with no crime figures. A pin in the sea or outside the UK shows none.

Countries

Every country page gives one figure that can be compared across borders: intentional homicides per 100,000 people, from the World Bank, which republishes the UN Office on Drugs and Crime (UNODC) series. Taiwan, which the World Bank list leaves out, uses the homicide victims counted by its National Police Agency. Where a country has no World Bank figure from the last 5 years (much of Africa), its page uses the World Health Organization's Global Health Estimates instead: a modelled estimate of the homicide rate, not a count, and says so. Police in different countries record theft, assault and other crimes so differently that their totals cannot be compared; a homicide is recorded in much the same way everywhere. Each page shows the latest year with a figure, the change over about 10 years where there is a figure from 8 to 12 years earlier, the world median, and the country's place among its neighbours (the UN sub-region, for example South America). Only countries with a figure from the last 5 years are ranked, so an old figure is never set against a recent one. Countries get a 1 to 10 score only against the whole world (below). A country with no figure in the last 10 years still has a page, which says so. Pages for 38 European countries also list crime recorded by the police by type (assault, sexual violence, robbery, burglary, theft, drugs, fraud, corruption and more), from Eurostat (dataset crim_off_cat), the latest year for each type. Those figures are shown for each country on its own and never compared or ranked: each country defines and records them in its own way. Country outlines come from Natural Earth, and place names and postcodes outside the UK in search from GeoNames (CC BY 4.0). A street address that is not in those lists is looked up by Photon (komoot), on OpenStreetMap data (ODbL), straight from your browser.

Places and neighbourhoods outside the UK

Where a country publishes crime by region, town or police agency, each place gets a Safety score from 1 to 10 compared only with other places in the same country (or the same state, where each state counts crime its own way, as in Australia), never with places abroad. Crimes are weighted by type with one table for every country: homicide 100, sexual offences 20, violence 10, robbery 8, theft and burglary 2, drugs 2, damage and public order 1, other 1. Where a source publishes only its total of recorded crime, with no types, places are ranked by that total per 100,000 residents, unweighted, and only against places from the same source. The weighted total per 100,000 residents is ranked in tenths: 10 is the lowest tenth. A place is ranked only with a full year of figures, at least 30 comparable places, and enough residents for a rate to mean something; the page says when one was left out and why. Figures the source does not publish are shown as not published, never as zero.

A few cities also have neighbourhood scores, made the same way from each city's own small areas (New York neighbourhood tabulation areas, Dutch buurten, Tokyo chōme, Sydney suburbs, Buenos Aires barrios), with residents from each country's census. A pin there shares crimes and residents out by how much of each neighbourhood is inside the circle, as in the UK. To show how far a neighbourhood score can be trusted we test it against the year before: each neighbourhood page gives the share of neighbourhoods whose score moved by at most 1 between the two years. There is no back-test like the UK's, because these sources do not go down to single streets.

Compared with the world

Each scored place also gets a second Safety score from 1 to 10, which compares it with the whole world rather than only its own country. Local crime cannot be compared directly across borders, because police in each country count it differently, so this score rests on the one figure counted in much the same way almost everywhere: each country's intentional homicide rate per 100,000 people (the figure on the country pages). A place keeps its position within its country (how much more or less crime it records than the country's typical place), and that position is carried onto the world scale through its country's homicide rate, so a country's typical place gets the same world score as the country itself, which is the colour the country takes zoomed out. The world's tenths are weighted by population: each country counts by its number of residents, not by its number of places, so a country with tens of thousands of small communes does not outweigh its neighbours, and the most populous countries, such as China and India, weigh most. Places ranked within a city or state that has no figure of its own above them take their country's level, so differences between the cities or states of one country are lost there. Where a country publishes figures for very few places, those places take a level close to their country's. Countries whose latest homicide figure is more than 5 years older than the newest figure in the data are not compared and show as not comparable. Until 25 October 2026 this score compared each place with its own continent; it now uses one scale for every country, so a country that is calm for its region but has several times Europe's murder rate no longer shows as very low crime. The score cannot see crime that a homicide rate does not reflect: a country with few murders but a lot of theft looks calmer than it may be.

Living outside the UK

In European countries the pin, place pages and country pages also give a Living score from 1 to 10: how close a spot is to schools, public transport, a doctor and a hospital, and a supermarket, each scored from 1 to 10 by distance cut-offs in metres. It is kept separate from the Safety score and never blended with it. The places come from OpenStreetMap, which volunteers map, so where it has nothing for a theme in a country that theme is left out and the score is made from the rest. Schools count by distance only: outside the UK there are no inspection grades to use, so nothing here says whether a school is good. Each place is scored at one point and ranked in tenths against places of the same kind in its own country (10 is the best tenth); a region or country gets the average of its places. Where a country has too few places to rank, the score out of 10 is shown unranked. Points of interest: © OpenStreetMap contributors, under the Open Database Licence.

Where the data comes from, and how recent it is

• Police recorded crimeSeptember 2016 to August 2026. Some forces stopped or paused publishing, so the dates differ by area; the page says so where it matters.
• Detailed crime types and knife crime (Home Office)Financial year 2025/26, published quarterly.
• Ofsted inspections and school details (England)Newest version published when this release was built, 26 October 2026.
• GP surgeries and CQC ratings (England)Newest version published when this release was built, 26 October 2026.
• Railway stations, tram stops and bus stops (NaPTAN, Translink)Newest version published when this release was built, 26 October 2026.
• Supermarkets (Geolytix)Newest version published when this release was built, 26 October 2026.
• Hospitals and some GP locations (OpenStreetMap)Newest version published when this release was built, 26 October 2026.
• Neighbourhood boundaries and population (ONS, NISRA)2021 boundaries with mid-2024 population estimates. Postcode lookup: ONS National Statistics Postcode Lookup, Newest version published when this release was built, 26 October 2026.
• Scotland: Data Zones, schools and council recorded crime (Scottish Government), population (National Records of Scotland), GPs and A&E (Public Health Scotland)2022 Data Zones with mid-2024 population estimates. Newest version published when this release was built, 26 October 2026.
• Place and street names (OS Open Names)Newest version published when this release was built, 26 October 2026.

The figures are rebuilt once a month; the current release was built on 26 October 2026.

What it cannot tell you

• Greater Manchester Police has published nothing since June 2019, so Greater Manchester shows crime from 2016 to June 2019 with a notice. British Transport Police stopped in February 2025.
• In Scotland there is no Safety score yet; Living covers it. Police Scotland does not publish street-level crime in the same open form as police.uk. Scottish council pages show the Scottish Government's recorded crime by group, with the share cleared up.
• Schools scores and CQC ratings are England only. Wales and Northern Ireland have no schools score.
• Crime is recorded where it happens but measured per resident, so busy city and town centres, where few people live and many come to shop, work and go out, show rates far higher than the risk to someone living there. We treat a neighbourhood as a busy centre when it records at least 400 crimes per 1,000 residents a year and at least 30% of them are theft from shops or from people, or when it records more crimes than it has residents. Busy centres get no Safety score, because compared with residential areas the number would mislead, and they are left out when every other neighbourhood is ranked or blended. Their real crime figures are still shown, and clicking a centre on the map explains why it is not scored. The same applies to a postcode district or a map circle where half or more of the residents live in busy centres.
• Where a radius cuts through a neighbourhood, residents are assumed to be spread evenly across it, so those figures are estimates.
• Recorded crime depends on what people report and how forces record it. Rates are not directly comparable between years when recording practice changed.

Live cameras

The camera dots on the world map show public traffic and city cameras. Each picture comes straight from the operator, or through our server when the operator needs a key or blocks direct loading; we do not keep the pictures. Sources:

• Powered by TfL Open Data. Contains OS data © Crown copyright and database rights
• Servei Català de Trànsit; Ajuntament de Barcelona; Ajuntament de Terrassa
• Fintraffic / digitraffic.fi, license CC BY 4.0
• Contains data under the Norwegian licence for Open Government data (NLOD) distributed by Statens vegvesen
• City of Tallinn, ristmikud.tallinn.ee
• DGT (CC BY)
• Ayuntamiento de Madrid (CC BY 4.0)
• CITA, Administration des ponts et chaussées (CC0)
• Autostrada del Brennero (CC0), via Open Data Hub
• Département de Loire-Atlantique (Licence Ouverte 2.0)
• Vegagerðin
• Ayuntamiento de Málaga
• Ayuntamiento de Vitoria-Gasteiz
• MeteoGalicia, Xunta de Galicia
• DfI Traffic Information and Control Centre (TrafficwatchNI)
• Hong Kong Transport Department / DATA.GOV.HK
• Land Transport Authority / data.gov.sg (Singapore Open Data Licence)
• Live Traffic NSW, Transport for NSW (CC BY 4.0)
• Contains information licensed under the Open Government Licence – Toronto
• Contains information licensed under the Open Government Licence – City of Calgary
• DriveBC. Contains information licensed under the Open Government Licence – British Columbia
• Waka Kotahi NZ Transport Agency
• Ontario 511, Ministry of Transportation of Ontario
• Georgia 511, Georgia Department of Transportation
• Ohio Department of Transportation, OHGO
• Caltrans
• YouTube

Attribution and licences

• Crime: street-level data from data.police.uk, published under the Open Government Licence v3.0.
• Detailed crime types and knife crime: Home Office, police recorded crime open data tables (Community Safety Partnership tables; offences involving knives or sharp instruments), under the Open Government Licence v3.0. Council-to-partnership lookup: Office for National Statistics, under the Open Government Licence v3.0.
• Crime severity weights: Office for National Statistics, Crime Severity Score (year ending March 2025), and Home Office police recorded crime open data (financial year 2025/26, used to average the weights inside each police.uk category). Contains public sector information licensed under the Open Government Licence v3.0.
• Neighbourhood boundaries and population: Office for National Statistics (LSOA 2021 boundaries, mid-2024 estimates) and NISRA (Data Zones 2021, mid-2024 estimates), under the Open Government Licence v3.0.
• Scotland: Scottish Government (Data Zone 2022 boundaries, school locations, recorded crime and crime clear-up rates), National Records of Scotland (small-area population estimates, mid-2024) and Public Health Scotland (GP practices and A&E sites), under the Open Government Licence v3.0.
• Postcode lookup: ONS National Statistics Postcode Lookup. Source: Office for National Statistics licensed under the Open Government Licence v3.0.
• Contains OS data © Crown copyright and database right 2026.
• Contains Royal Mail data © Royal Mail copyright and database right 2026.
• Contains Ordnance Survey of Northern Ireland information licensed under the Open Government Licence v3.0.
• Supermarkets: Geolytix UK Retail Points, published under the Geolytix Open Data Licence. Contains public sector information licensed under the Open Government Licence v4.0; Contains Ordnance Survey data © Crown copyright and database right 2021; Contains Royal Mail data © Royal Mail copyright and database right 2021; Contains National Statistics data © Crown copyright and database right 2021.
• Bus stops, railway stations and tram stops in Great Britain: DfT National Public Transport Access Nodes (NaPTAN), under the Open Government Licence v3.0.
• Northern Ireland bus stops and railway stations: Translink, via OpenDataNI, under the Open Government Licence v3.0.
• Schools: Get Information About Schools (GIAS), Department for Education, and Ofsted inspection outcomes, both under the Open Government Licence v3.0.
• GP surgeries: NHS England Organisation Data Service (ODS), under the Open Government Licence v3.0. GP ratings: Care Quality Commission (CQC), under the Open Government Licence v3.0.
• Hospitals and some GP locations © OpenStreetMap contributors, available under the Open Database Licence (ODbL), via Geofabrik extracts.
• Base map: OpenFreeMap, using OpenMapTiles and © OpenStreetMap contributors.
• Living (schools, transport, doctors and hospitals, supermarkets outside the UK): © OpenStreetMap contributors, available under the Open Database Licence, via Geofabrik extracts.

The Open Government Licence v3.0 is available at nationalarchives.gov.uk.

Accessibility

Crime Rates Map is built to meet WCAG 2.2 level AA: it works with a keyboard and a screen reader, at any zoom, on any screen size. If anything is hard to use, tell us at dev@rafaelcardoso.co.uk and we will fix it.