Methodology

Source

The Department for Transport publishes vehicle licensing statistics as a set of data files on its vehicle licensing statistics data files page. Every figure on these pages comes from one of them, VEH0160 (UK): first registrations of vehicles by body type, make, generic model, model and fuel, with one column per quarter, covering Great Britain and Northern Ireland from 2014 Q3 onwards.

The file is a single CSV of about ten megabytes. Its download address changes with every release — the DfT re-uploads it under a new media identifier — so the pipeline never stores a fixed URL: on each run it fetches the statistics page and resolves the link whose file name is df_VEH0160_UK.csv. The file is archived before it is parsed, so every published figure can be traced back to the exact release it came from.

File extracted: 2026-09-07. Latest quarter: 2026-Q1, with data through 2026-03-31. The site shows the latest 12 quarters, three full years.

Cadence and lag

The United Kingdom is the only quarterly market on auto-stats.eu. The DfT publishes a quarter roughly three and a half months after it ends (quarterly key tables around July, September and December, the full annual set in April or May), and the refresh runs every week so a new quarter appears within days of its release. Nothing on these pages is a calendar month, and no figure is provisional in the sense used on the monthly markets — the DfT publishes a quarter only once it is complete, and then keeps improving it (see Restatements below).

What is counted

A first registration is a car registered for the first time in the United Kingdom. Only body type Cars is counted: passenger vehicles with up to eight seats. Taxis, light goods vehicles, motorcycles and buses are separate body types in the same file and are excluded here.

The DfT publishes no split between brand-new cars and used imports, so both are counted together, and used imports — roughly 2.5 % of the total — are included in every figure on these pages. That also means there is no import origin, no age at import and no CO₂ at make or model level for the United Kingdom. The SMMT's widely quoted monthly new-car registration totals count new cars only and therefore run about 2.5 % below the figures here (2,020,520 against 2,071,394 for 2025); the two are measuring different things and should not be compared directly.

Generic model versus model

VEH0160 carries two model columns. GenModel is the DfT's generic model — the grouping shown on these pages — and Model is the trim-level description, of which there are tens of thousands. Trim rows are summed into their generic model, so a "Ford Puma" row here is every Puma trim in the quarter.

• Rows whose generic model is <MAKE> MODEL MISSING — about 1 % of car volume — carry no model at all. They are included in the brand totals and in the KPI totals, but they appear in no model row, so the model table adds up to slightly less than the total.
• Newly coded generic models sometimes still carry trim text (JAECOO 7 LUXURY 4X4 AUTO); the alias tables in the repository fold those back onto the model.
• A few DfT groupings are wider than the marketing name — BMW IX holds the iX, iX1 and iX2 — and are left exactly as the DfT publishes them rather than being split on a guess.

Fuel groups

The DfT's fuel labels are mapped to six groups:

• Petrol and diesel: cars with no electric assistance.
• Hybrid: the DfT's "Hybrid electric (petrol)" and "Hybrid electric (diesel)". This group includes mild hybrids (MHEV) alongside full hybrids — the DfT does not separate them — so the UK hybrid share is much higher than, and not comparable with, a market that counts full hybrids only.
• Plug-in hybrid: the DfT's plug-in hybrids, petrol and diesel, plus range-extended electric cars.
• Battery electric: cars with no combustion engine.
• Other: fuel cell electric, gas and the DfT's "other fuel types" residual, together well under 1 % of registrations.

Restatements

The DfT states that its make and model coding is improved iteratively with each release, so figures for earlier quarters move between releases. This site does not try to freeze them: the newest file is authoritative, and the whole 12-quarter window is rebuilt and republished on every refresh. A quarter whose total moves by more than 1 % against the previous run is listed in the refresh pull request so the change is visible, and a move above 10 % stops the run for a human to look at.

KPI definitions

First registrations

All cars registered for the first time in the United Kingdom in the quarter, from the DfT's VEH0160 data file. DfT does not separate brand-new cars from used imports, so both are counted here (used imports are roughly 2.5 % of the total).

BEV share / PHEV share

Battery-electric cars as a share of all first registrations in the quarter. PHEV adds DfT's plug-in hybrids and range-extended electric cars.

Hybrid share

DfT's "hybrid electric" group as a share of first registrations: full hybrids and mild hybrids (MHEV) together, petrol and diesel. Plug-in hybrids are counted separately.

QoQ / YoY / YTD

Change against the previous quarter, against the same quarter a year earlier, and the calendar year to date with its change against the same quarters a year earlier. Percentages are shown only when the comparison quarter has at least 30 registrations, or 10 on the electric tables.

Checks

Every refresh has to pass the same gates before anything is published. The identity gate recomputes the latest quarter's total and each fuel group straight from the raw file and requires them to match the exported figures exactly. The acceptance gate reviews the largest 40 makes for coverage of at least 97 % of the last four quarters, for sibling makes that share a prefix and should be one brand, and for single-letter model names other than the Mercedes-Benz classes and Honda's e. The sanity gate rejects an export whose latest quarter is less than half the previous one, whose marts are empty, or whose quarter series has a gap since 2023 Q1. The fuel gate fails when more than 2 % of a quarter's registrations land in the "other" group. The restatement gate is described above. Finally a consistency audit re-derives every share, rank, QoQ, YoY and year-to-date figure from the exported JSON files and fails on any mismatch with what the pages show.

Caveats

• Brands with fewer than 500 first registrations over the 12 quarters have no page of their own and appear as plain text in the tables.
• DfT's hybrid group includes mild hybrids, so the hybrid share is not comparable with markets that only count full hybrids.
• Counts are DfT's first registrations of cars (body type Cars) in the United Kingdom, new cars and used imports together; DfT publishes no split between them.
• Registration is not sale: pre-registered dealer stock, demonstrators, rental and leasing fleets are all first registrations.
• Percentages are shown only when the comparison quarter has at least 30 registrations.

Licence and data files

Contains public sector information from the Department for Transport and the DVLA licensed under the Open Government Licence v3.0. Site content and derived data are free to reuse under CC BY 4.0 with attribution to "auto-stats.eu (data: DfT)".

Every page here is generated from JSON under /data/uk/: meta.json, periods.json, YYYY-Qn.json (with kpis, models[], makes[], bev_models[] and bev_makes[]), makes.json and makes/<slug>.json.