Towns of England & Wales

Every actual town and city in England and Wales, 1,428 of them, drawn from the real extent of the buildings rather than from any council boundary. Each circle is a settlement, sized by how many people live in it and coloured by whichever measure you pick.

This is the companion to the local authority heat map, and it answers the complaint that map invites: councils are not places. Cornwall is one authority covering half a county; Birmingham is one authority covering a city of a million. Here the unit is the built-up area, which the ONS draws around the bricks themselves, so Reading is the town of Reading and Blackpool is the seaside resort, not the district that happens to contain it. Zoom in and the circles give way to each town’s actual footprint.

The trade is that far less is published at this scale. Seven of the nine measures come from Census 2021, and the other two are recorded crime, joined from police street-level data. There are still no pay, house price, life expectancy or deprivation figures for towns, because nobody produces them. It is England and Wales only, for the same reason the other map stops at the Welsh border going north.

Loading the map…

Every town, ranked

The map’s data as a table, highest first, for the measure selected above. Type to find a town.

What a “town” is here

A built-up area is the ONS’s outline of continuous development: the land actually covered by buildings, with the fields left out. That makes it the closest official thing to a town, and quite unlike a council area, which is drawn for administration and often bundles a town together with a large tract of countryside or with its neighbours.

England and Wales contain 7,713 built-up areas, but most are hamlets. This map keeps the 1,428 with 5,000 residents or more, which is the conventional floor for calling a settlement a town, and which between them house about 79% of the population of England and Wales. The rest of the population lives in villages, hamlets and open country that no town map can show.

The big conurbations are split into their constituent named places rather than lumped into one blob. There is no entry called London: it appears as Croydon, Ealing, Camden, Newham and the rest. Greater Manchester likewise comes through as Manchester, Salford, Stockport, Oldham and their neighbours. That is the ONS’s choice, not this page’s, and it is a reasonable one, since those are real named places, but it means the biggest circle on the map is Birmingham rather than London.

Where the numbers come from

Nothing is published against built-up areas, so every figure here is summed up from output areas, the smallest census geography, at roughly 130 households each. A build script pulls the Census 2021 tables for all 188,880 output areas in England and Wales from NOMIS, then adds them into towns using the ONS’s own output-area-to-built-up-area lookup. Every number is a real census count, added up. None of it is modelled, interpolated or apportioned.

The one wrinkle is that the lookup is best fit: an output area is assigned whole to a single built-up area, so a town’s edge is accurate to roughly a block rather than to the individual house. The boundaries are generalised to about 110 metres for the web, which is invisible at the zoom you would look at a single town from but does round off the frayed edges of a town’s outline.

A caveat on the crime figures

The two crime measures come from the street-level records every police force publishes to data.police.uk: 5.85 million offences over twelve months, each carrying the small statistical area it happened in, joined to towns by that code rather than by its coordinates. The coordinates in that dataset are deliberately fuzzed to an anonymised point, so using them would invent precision the data doesn’t have.

Greater Manchester is blank, and that is not an error. Greater Manchester Police does not publish to data.police.uk at all: there is not one file from them in the entire archive. Manchester, Salford, Bolton, Oldham and their neighbours are therefore left unshaded and out of the rankings rather than shown near zero, which is what counting the handful of cross-border offences from neighbouring forces would have produced. That is 74 of the 1,428 towns.

Two things to hold on to when reading the rest. Recorded crime is what was reported to and recorded by the police, which is not the same as what happened, and willingness to report varies from place to place. And the rate is per 1,000 residents, so a town that fills up daily with people who live elsewhere carries their offences without their heads: the City of London tops the table by a mile on 8,000 residents, and seaside and retail towns sit far higher than the people who sleep there would suggest. Burglary is the steadier of the two, being mostly a crime against premises.

A caveat on the data

Census 2021 was taken in March 2021, during a period of pandemic restrictions. That is known to have moved some people around, particularly students, who were counted at their term-time address even where they had gone home. University towns are the place to be most careful: their population, age profile and qualification figures all lean on where students were recorded. The data is also now some years old, and towns that have built out since will read low.

Circle size is proportional to the square root of population, so that area on screen tracks population rather than radius doing it. Even so, a circle is a rough guide: use the number, not its size, when the difference matters.

Some of the figures in the charts and tables on this page were compiled with the help of AI tools and may contain errors or be out of date. They are shared in good faith for general interest only (not as professional, financial, investment or purchasing advice) and should be checked against the cited primary sources before you rely on them.