uHouse methodology
How uHouse currently builds each figure — what it is based on, how it is computed and where its limits are. Source attributions are on the Sources & copyright page (/sources).
Scope
uHouse covers residential property in England and Wales: every home that appears in the official register of residential sales since 1995. Scotland and Northern Ireland keep separate registers and are not covered.
Homes and addresses
Each sale record carries a postcode, a secondary address (flat or unit), a primary address (house number or name) and a street. A home is the set of sales sharing the same normalised postcode, flat, house number or name and street — upper-cased, punctuation removed, spacing collapsed. Records without a postcode fall back to the district so that they do not merge across the country.
Where the sales register publishes the Unique Property Reference Number (UPRN) of a sale, the home carries it.
Location
A home is placed at its UPRN's official point when its UPRN is known (from the sale or from a matched energy certificate); otherwise at the centre of its postcode. Planning constraints are only shown for homes placed at their own UPRN point, because a postcode centre is not the building.
Energy certificates
The latest Energy Performance Certificate for each home is matched by UPRN first; failing that, by postcode plus normalised address, and only when exactly one certificate address fits. Floor area, rooms, construction age, energy rating and main fuel come from that certificate.
Improvement steps are the assessor's recommendations on that certificate, in the assessor's order, with the indicative cost range it states; the rating the home could reach is the certificate's own potential rating. Costs are the assessor's indicative figures, not quotes.
Recent sales nearby
Standard (category A) sales in the same postcode over the last five years and on the same street, in the same town, over the last three. Category B transactions — repossessions, buy-to-let company purchases and other transfers that do not reflect a market price — are left out. Price per square metre divides the sale price by the matched certificate's floor area.
Local price trend
The official house price index for the home's local authority, monthly over ten years, for all homes and for the home's own type. The index price is mix-adjusted — the price of a typical home in the area — not the average of that month's sales. The five-year change compares the latest month with the same month five years earlier.
Schools
The three nearest open primary and secondary schools within 5 km (straight-line), from the official school register. The Ofsted line shows the most recent outcome that states a judgement: a report-card inspection (from November 2025) with its area judgements; a graded inspection with its overall grade; a graded inspection from September 2024 to November 2025, which gave no overall grade, with its quality-of-education judgement; or a later ungraded inspection with the grade it confirmed.
Transport
The nearest rail station and Underground, metro or tram stop within 5 km and the nearest bus stop within 1 km, from the national public-transport stop register, active stops only; plus the number of bus stops within 400 m. Distances are straight-line.
Crime
Police-recorded street-level crime, counted for each neighbourhood (Lower Layer Super Output Area, about 1,500 residents) over the latest twelve published months, by type. Larger areas — census neighbourhoods of about 8,000 residents, councils, England and Wales — add up their neighbourhoods, so every rate is weighted by population.
Crime rate: recorded crimes a year over the latest 24 published months per 1,000 residents, each year against the people living there that year — the official mid-year population estimate plus the new homes built since, so fast-growing areas are not judged against an old head count. Two years rather than one, because a neighbourhood records few enough crimes for a single year to swing by chance. Crimes the police place at a non-residential venue — a supermarket, shopping area, petrol station, nightclub, hospital, college, theatre, bus or coach station, conference centre, police station or prison — are left out of the rate and shown apart: they say more about the venue than about living nearby. Anti-social behaviour (noise, nuisance, rowdiness, littering) is not part of the rate and is shown separately, with its count and per 1,000 residents: it is not part of recorded-crime statistics, how it is recorded varies a lot between police forces and over time, and adding it in would hide real differences in crime between areas. It still counts in full in the safety score. Bands: under 50 very safe, 50–80 normal, 80–120 above average, 120–150 clearly high, 150 or more high risk. The England and Wales rate sits above that of the middle neighbourhood, because a small number of neighbourhoods record a very large number of crimes.
Safety score, 0 to 100 (100 = safest): each crime is weighted by how much it bears on the people who live there — violence and sexual offences, burglary, vehicle crime and anti-social behaviour count in full; robbery, possession of weapons and criminal damage and arson 0.6; public order, drugs, bicycle theft and other crime 0.3; other theft and theft from the person 0.15; shoplifting not at all, so a shopping street is not ranked by its shops' losses. The weighted total per 1,000 residents, on the same footing as the rate (24 months, venues left out), is placed among all neighbourhoods in England and Wales: the score is the share of neighbourhoods with a higher weighted rate. Larger areas are placed on the same neighbourhood scale. The score's label uses the same five bands, cut where they hold the same shares of neighbourhoods as the rate bands (score 53 or more very safe, 27–52 normal, 11–26 above average, 7–10 clearly high, under 7 high risk).
Trend: the latest twelve months against the twelve before, same definition. More than 10% up is rising, more than 10% down is falling, anything between is stable; with fewer than 40 crimes over the two years together no direction is given. The monthly chart shows the latest 36 months.
Busy centres: a rate per resident overstates the risk to residents where few people live but many visit — town and city centres, shopping streets, stations, airports. Leaving out the venues' crimes covers part of this; an area is still flagged when its latest twelve months' crimes, venues included, are at least 250 per 1,000 residents (about three times the national rate) and either a quarter or more of its crimes are shoplifting or theft from the person, or fewer than 500 people live per square kilometre; or when its rate is 500 per 1,000 or more.
Gaps in the published data: a police force sometimes publishes no street-level data, or only part of it, for a month, and may add it later. A month counts as missing for a council when it holds fewer than half the crimes of the median month in the two years before it. With one or two such months among the latest twelve, the remaining months are scaled up to a year; with three or more, the figures use the most recent twelve complete months instead, and the page names them. The trend then compares the same months a year apart. A council without twelve complete months in the last 36 is flagged as possibly reading low and is left out of rankings.
Crime on trains and at stations: crimes recorded by British Transport Police — on trains, the London Underground and at railway stations — are left out, for every area and every month. They are missing from the published data for most months since March 2025, so counting only the months that exist would make neighbourhoods around stations read low and look as if crime there were falling.
Harassment is not a separate category in police-recorded crime: it is counted within violence and sexual offences and within public order. Where a police force publishes no street-level data the area shows no rate or score, and the national figures leave it out. Locations are anonymised to a fixed set of map points before publication, so crimes near a neighbourhood's edge can be counted in the neighbour.
Flood zones
The official flood zones for planning, rivers and the sea. Outlines are simplified to 2 m before lookup; a home inside Zone 3 is reported as Zone 3, inside Zone 2 as Zone 2, otherwise Zone 1. England only.
Neighbourhood profile
Deprivation deciles come from the official 2025 deprivation index of each nation, England's for English neighbourhoods and Wales' for Welsh ones: decile 1 is the most deprived tenth of neighbourhoods in that nation, 10 the least. The two indices rank within their own nation and use different domains, so a decile in England and one in Wales are not directly comparable. Census 2021 shares — tenure, accommodation type, age, qualifications, car ownership — are computed from the published counts and shown beside the England and Wales share.
Local rents
The official private rents index for the home's local authority: the average monthly rent across all private tenancies, new and existing, overall, by number of bedrooms and by property type, beside the figure for England or Wales. It describes the local rental market, not the rent this home would achieve.
Everyday essentials
The nearest public park or garden and play space are measured to their nearest entrance; green space within 1 km excludes religious grounds, cemeteries and golf courses. Supermarkets are grocery stores of the main chains with their size band; GP practices are active practices on the official register, placed at their postcode; hospitals are NHS sites named as a hospital or infirmary. All distances are straight-line.
Broadband and mobile
Official coverage data: for the home's postcode, the share of premises that can get superfast (30 Mbit/s or more), ultrafast (300 Mbit/s or more) and gigabit-capable broadband, and the share below the universal service obligation (a decent 10 Mbit/s service); for its local authority, the share of premises with good outdoor 4G / 5G signal from all four networks and from at least one. This is availability reported by providers, not a speed test.
Environmental risks
Flooding from rivers and the sea uses the official national flood risk assessment: the band of the home's own address point when its UPRN is known, otherwise how many homes in its postcode fall in each band. Flooding from surface water uses the same assessment's surface water areas, checked at the home's address point. In Wales both come from the Welsh official flood risk assessment (rivers, sea, and surface water and small watercourses), checked at the address point; its bands allow for flood defences and it has no "very low" band.
Road and rail noise are the official strategic noise maps for 2022, for England and for Wales, read at the home's address point: Lden is a day–evening–night average with evening and night weighted up; Lnight is the night-time level. Levels below 50 dB Lden are not shown.
Radon is the official indicative radon atlas: for each 1 km square, the highest estimated share of homes above the radon action level within it. Air quality is the official modelled background annual mean for the current year on a 1 km grid, away from roadside peaks. Coal mining risk areas are the official development risk areas.
All of these are model- or area-based indications, not a survey of the property.
Ground stability
Ground stability is an area indicator from the national geological overview of six natural ground hazards: clay shrink–swell, landslides, compressible ground, collapsible deposits, running sand and soluble rocks. Great Britain is divided into hexagons with 5 km sides, about 65 km² each, and each hexagon has one class per hazard — low, moderate or significant — generalised from the detailed 1:50,000 susceptibility map. For landslides the class is the highest found in the hexagon; for the other five it is the class covering the largest share of it, and where smaller significant areas occur inside a lower-rated hexagon the page says so.
We show the hexagon that contains the home's point. It describes the surrounding area, not the ground under one property, and it covers natural ground conditions only (not mining or made ground, apart from soft man-made ground in the compressible layer). A surveyor can advise on a specific home.
Planning constraints
Conservation areas, Article 4 direction areas and tree preservation zones that contain the home's point, and the listed building it sits in (a listed-building outline containing the point, or a list entry within 10 m), from the official planning registers. Councils are still completing this data.
Sun and shadow
Building outlines are OpenStreetMap's, with the Ordnance Survey OS OpenMap Local buildings OpenStreetMap lacks. Heights and tree crowns come from 1 m LiDAR: the Environment Agency's national composite in England and the Welsh Government's 2020–23 survey in Wales. A building's drawn height is the 90th percentile of surface minus ground height inside its outline; shadows are cast from a lower, flat block (the median), so a pitched roof does not shade as if it were flat at its ridge. A roof of more than one height is split into parts. Buildings put up since a survey are estimated from OpenStreetMap floor counts, the home type or the footprint size; an extension built since counts as one storey, or two where an approved planning application describes a two-storey extension.
Every 10 minutes of the chosen day the sun's position is worked out and a line traced towards it: buildings and trees within 450 m block it, and beyond that the skyline from a 30 m elevation model out to 20 km. Walls are checked 1 m out from the wall at window height. Hours assume a clear sky and are estimates for comparing homes, not a survey of a particular window. The back garden is taken as the side away from the nearest named street in the home's address. Scotland and Northern Ireland are not covered.