UK Service History Coverage Report 2026: Brand by Brand

Why we ran this research

When you buy a used car in the UK, "full service history" is one of the most over-used phrases in the classified ads. But how often is it actually true? DVLA publishes registration data and the DVSA publishes MOT data, but nobody publishes how often manufacturers' own digital service systems actually return records when you query them by registration plate — or how often the mileage on those records doesn't add up.

We can. Across the last 24 months, FindServiceHistory has run thousands of dealership service-record and MOT lookups for UK car buyers and motor traders. We aggregated this dataset (with all personal data stripped out) to answer two questions: which UK car brands actually have retrievable digital service records, and how often does a car's recorded mileage go backwards?

Methodology

This report is based on 3,626 anonymised UK vehicle lookups conducted between August 2024 and August 2026, encompassing 12,917 individual dated workshop entries and more than 20,000 MOT test records. Coverage is defined as the percentage of lookups where the manufacturer's digital service record system returned at least one logged entry. Note that wording: an entry is a workshop visit, and only about half of them are a scheduled service — which is why the service-interval figures further down use a deliberately narrower definition than the coverage figures.

A brand's hit rate is not a measure of its feed. Our data source classifies each manufacturer into one of three contracted tiers: full ("we expect to receive data on any recorded services"), limited ("data in a minimal proportion of cases") and workshop remarks (recalls and warranty work only — services are not included). Those are three different things, not one scale, so a low hit rate on a full-coverage brand says nothing about the feed: it is a property of the population being looked up. Our own book is UK retail and van trade, where a great deal of servicing happens outside the franchised network and therefore never reaches the manufacturer's system at all.

Sensitivity check. Three brands returned zero records across their entire sample and are not contracted to return services in the first place — Hyundai (workshop remarks, n=72), Suzuki (limited, n=31) and MG (n=34). For those, a blank result carries no information about the car. Excluding all 137 of them, the headline moves from 36.5% to 34.0% with no service history. We publish the all-brands figure as the headline because it is what a buyer actually experiences when they run a check, and the feed-adjusted figure here so the difference is visible rather than argued about. We do not exclude brands merely for having a low hit rate — that would condition the result on the outcome and would inflate the headline by around twelve points.

Brand-level coverage results are reported only where we have at least 30 lookups for that manufacturer (24 brands meet this threshold). MOT rankings require at least 50 MOT tests on file (32 brands). Service intervals count only entries whose itemised work confirms scheduled maintenance, and require at least 30 measurable gaps between consecutive such visits on the same vehicle (12 brands). Mileage- anomaly rates are shown per brand only where we have 50 or more vehicles with two or more MOT readings. No vehicle, owner, or registration is identifiable in any figure below.

One correction we have applied to our own data. For three days in July 2026 our service-history data provider stopped answering queries, and 85 lookups were recorded as "no records" when in fact nothing had been asked successfully. We re-ran all 85: 46 did have service history and are counted here as such (using their re-verified records), 20 were confirmed genuinely empty, and the 19 that could not be re-verified are excluded from the sample entirely rather than guessed at. Reporting a supplier outage as an absence of service records would have overstated the headline figure by roughly 1.5 percentage points.

What "no service history" means — and doesn't. A blank result does not mean the car was never serviced. It means there is no digital record retrievable from the manufacturer's own service database. That happens for two very different reasons: the car genuinely wasn't serviced (or was serviced only at independent garages that never fed records back), or the manufacturer simply doesn't make its digital records readily available to the lookup systems the trade uses. Some brands — notably Ford and most Korean and Japanese marques — fall heavily into the second category. So a low coverage figure for those brands is largely a measure of data availability, not owner neglect. We flag this throughout, because it changes how you should read the brand table.

Headline finding: more than a third of used cars have no digital service trail

Across all 3,626 vehicles in the sample, 1,324 (36.5%) returned zero service records from the manufacturer's digital system. A buyer relying on a digital check would, in more than one case in three, find nothing.

The conventional wisdom in the UK motor trade is that the digital-service-history rate is well above 70% for cars under 10 years old. Among the scrutinised vehicles in our sample it's considerably lower — and, as the next section shows, the variation between brands is far wider than anyone had previously quantified.

Digital-record availability varies hugely by brand

Of the 24 brands with enough sample size to publish, the share of lookups that returned at least one digital service record:

The premium German brands sit at the top — a digital check on a used BMW, Audi or Mercedes returns records in the large majority of cases. At the other end, Ford (10.6%) and Hyundai (0% in our sample) return records rarely or never.

Read this as an availability map, not a neglect map. The gap is driven mainly by how openly each manufacturer's digital service data can be retrieved. BMW, Audi and Mercedes built centralised, queryable dealer records early; Ford and most Korean and Japanese brands either digitised later or don't expose those records to third-party lookups. So Hyundai's 0% almost certainly reflects that its records aren't accessible via the systems we query — not that no Hyundai in Britain gets serviced.

The practical implication for buyers: the absence of a digital service history is much weaker evidence of neglect on a used Ford, Hyundai or Toyota than on a used BMW. For the low-availability brands, ask the seller for paper records or independent-garage invoices. For a BMW or Audi, a missing digital trail is a genuine red flag worth walking away from.

Mileage discrepancies: how often the odometer goes backwards

Because the DVSA records a car's mileage at every MOT, a rolled- back odometer usually leaves a fingerprint: a later test showing fewer miles than an earlier one. We checked all 2,615 vehicles in the sample with two or more MOT readings for exactly this.

38 of them (1.2%, about 1 in 80) had a reading that dropped by more than 1,000 miles between tests. Where a discrepancy appeared, it was rarely small: the median drop was 16,634 miles and the mean 40,650 — the equivalent of hiding well over a year's driving. This is a deliberately conservative floor: a rollback done between MOTs, or below the 1,000-mile threshold, wouldn't be caught, so the true rate is higher.

That figure is worth dwelling on, because we have now measured it three times on successively larger samples and got 1.4%, then 1.4%, then 1.2%. Adding roughly 1,500 cars across those runs moved the headline by two-tenths of a percentage point — which is the strongest evidence we have that somewhere around 1 in 70 to 1 in 80 is the real order of magnitude for scrutinised used cars, not an artefact of a small sample.

Among brands with 50+ eligible cars, the anomaly rate was highest for:

A caveat worth stating plainly: not every backwards reading is fraud. A replaced instrument cluster, or a keying error at the test centre, can produce the same pattern — which is why we report a discrepancy, not confirmed clocking. But every one is worth questioning before you buy. You can check any car's MOT mileage trail for free with our car mileage & MOT history checker, and our guide explains how to check if a car has been clocked.

A digital service record is a log of workshop visits — not a log of services

This was the most useful thing we found, and it changes how anyone should read a digital service record. A manufacturer's system does not record "services". It records every occasion the car was booked into a franchised workshop, and only about half of those visits turn out to be a scheduled service. Classifying every dated entry in the sample by the work actually itemised against it:

So a car showing twelve entries has not necessarily been serviced twelve times. Some of those entries are a recall on a brake hose, a warranty claim on a control unit, an investigation into a dashboard warning light, or the pre-delivery inspection from the day the car was first handed over. Count the services, not the stamps — and read the itemised work, which is exactly why our reports print it rather than just a date and a mileage.

How much a dealer network logs varies enormously by brand, so this is not a uniform effect. Of BMW's 2,818 entries only 662 record scheduled maintenance (23%); for MINI it is 121 of 663 (18%). At the other extreme, 1,442 of Audi's 1,539 entries do (94%), and Volkswagen is 95%. Both patterns are honest record-keeping — they are simply different conventions — but it means the raw number of entries in a record is not comparable between badges.

How often are these cars actually serviced?

Restricted to visits with confirmed scheduled-maintenance work, and collapsing multiple entries sharing one date into a single visit, this is the typical (median) gap between services. The right-hand column gives the manufacturer's schedule — but read the caution below it first, because for most of these brands there is no single figure:

The striking result is how little variation there is. All twelve brands fall between 9.7 and 13.4 months, so the dramatic brand spread implied by counting every workshop entry simply isn't there.

Read that as a comparison between badges, not as a rate for the fleet. It is tempting to round it off to "cars get serviced about once a year", and that would be wrong. The figure measures the gap between consecutive services recorded at a franchised dealer, on cars that had at least two of them, and it discards any gap longer than four years as a data error. Work done at an independent garage is invisible to it. The 36.5% of cars with no digital record at all are outside it entirely, and a car serviced once and then abandoned contributes nothing, because a single service creates no gap to measure. Those filters remove most of the ways a car can go unserviced, so a result near twelve months is close to what the method is bound to produce. It is a fair way to compare one badge against another — every brand is filtered identically — and a poor way to estimate how often the cars on the road are actually maintained.

A warning about that right-hand column: a brand does not have "a" service interval. It varies by engine, by model year, and by which schedule the car was sold on — often by a factor of two within a single badge. Jaguar and Land Rover's petrol Ingenium engines run to 16,000 miles a year, while the diesels run up to two years and 21,000 miles. Mazda's petrols go 12,500 miles between services; the Skyactiv-D 2.2 diesel goes about 6,200. Kia asks 10,000 miles of a turbocharged petrol and 20,000 of a naturally-aspirated one. Hyundai's petrols are annual and its diesels biennial. BMW, MINI and Mercedes-Benz publish no fixed mileage at all — their cars compute the interval from how they are actually driven. Renault's own published guidance is simply that the schedule "varies by model and engine type". The column above is therefore a rough guide to the regime, not a specification for any individual car.

This is also why we do not rank brands by "service discipline": the schedules being measured against are not the same, and often not even the same within a brand. A Land Rover at 13.1 months looks like the laggard of the table until you notice that the diesels in it are on a two-year schedule and are therefore being serviced roughly twice as often as required. Ranking that car against a Porsche on a fixed annual schedule tells you nothing about either owner. Judge a car against its own schedule, not against another brand's — and not against its own badge's average.

Mileage is where the real differences sit, and they are mostly about how the cars are used. Porsche owners cover only about 5,000 miles between services and Vauxhall owners around 5,300 — far inside any of the distance limits that apply to them, because the time-based half of the schedule triggers first on low-mileage and short-trip cars. BMW owners cover 13,665 miles, which sits comfortably within the range a condition-based system produces in normal use — the car asked for the service when it was due, which is the system working as designed. The VW Group brands are the most interesting: Audi at 9,163 miles, Škoda at 9,849, SEAT at 10,257 and Volkswagen at 10,596 all cluster near the ~9,300-mile fixed interval and nowhere near the 19,000-mile Longlife one, which suggests most VW Group cars in the sample are on fixed servicing rather than the flexible schedule the brochures emphasise.

What is excluded, and why. A quarter of all entries came back with no itemised work at all. We drop those rather than assume they were services: counting them would pull BMW's median back to roughly 8.7 months and reintroduce the exact bias this measure removes. Dropping them stretches the measured gaps, so on that count the figures are conservative. Pulling the other way, we discard any gap shorter than a month or longer than four years as a data error, and cutting the long tail shortens the median — a car that sat for five years between dealer visits is not represented here at all. The two effects work in opposite directions and we have not tried to net them off, so treat these as a like-for-like comparison between brands rather than a precise figure for any of them. A brand needs at least 30 measurable service-to-service gaps to appear, which is why twelve brands qualify here against twenty-one in the coverage table. Toyota, Kia, Lexus, Citroën, Dacia and Peugeot are absent because their records in this sample contain almost no itemised service work to measure; Ford is absent for the opposite reason — its records are well itemised, but too few Fords returned records at all to clear the 30-gap bar and publish a reliable figure.

MOT advisories: which brands hold up between tests

Cross-referencing with MOT data shows which brands stay clean between tests. We measured average advisory items per MOT test across 32 brands with 50 or more tests on file. The extremes — cleanest first, then the highest-advisory brands:

Bentley, Porsche and Smart collect the fewest advisories per test; MG, Abarth and Nissan collect roughly twice as many. Advisory count and fail rate don't always move together — Dacia is among the cleanest on advisories (0.88) yet carries the highest fail rate in the set (23.9%), while Lexus posts the second-lowest fail rate (11.3%) on a mid-pack advisory count. The full 32-brand table, including first-fail mileages, is available to journalists and researchers on request (see below).

What this means for car buyers

Four practical takeaways if you're shopping for a used car:

1. Treat "FSH" in the listing as a claim, not a fact. More than a third of cars have no digital service history at all — even when sellers describe them as having one. A service history check verifies the claim before you put down a deposit.
2. Check the mileage trail first — it's free. Around 1 in 80 cars shows an odometer reading that went backwards, and where it does the average amount hidden is substantial. Our free mileage checker flags it in seconds.
3. Calibrate your expectations by brand. A missing digital trail on a 7-year-old BMW is a serious red flag (because 19 in 20 do have records). The same gap on a Ford, Hyundai or Toyota is common and far less telling — ask for paper or independent-garage invoices instead.
4. Count the services, not the entries. Only about half the entries in a digital record are scheduled maintenance — the rest are recalls, warranty jobs and fault investigations. Read the itemised work and check the gaps against that car's own schedule — which depends on its engine, not just its badge. A diesel Land Rover on a two-year interval and a petrol one on a twelve-month interval are the same car to a listing and completely different cars to a service book. Our free service schedule checker gives you the manufacturer's interval to measure against.

For a step-by-step guide, see our how to check service history before buying article.

What this means for motor traders

For dealers buying stock at auction or on part-exchange:

• Premium-brand stock without a digital service history should be discounted harder. A BMW or Audi without records is statistically more likely to have a genuine maintenance gap than a Ford without records, where the blank usually just reflects data availability. Price the risk by brand.
• Always check the mileage trail on appraisal. A backwards MOT reading is the cheapest fraud signal to catch and the most expensive to miss on a part-exchange.
• Budget prep time for high-advisory brands. MG, Abarth, Nissan, Mazda and Suzuki all average well over 1.3 advisories per MOT — factor that into retail prep.

Dealers running 50+ checks a month can run lookups in bulk through our trade dashboard — or pay-as-you-go at £4.99 per check. See our dealer guide for how this fits into a buying workflow.

Caveats and limitations

Honest caveats, because nobody benefits from inflated claims:

• Coverage measures data availability, not maintenance quality. A car dutifully serviced at an independent specialist for 10 years returns zero records on a manufacturer digital check. The 36.5% no-history figure — and every brand figure — includes these cars.
• Selection bias: our sample is people who chose to run a check — pre-purchase buyers, traders evaluating part-exchanges, and owners recovering lost records. It is not a random sample of the UK car parc; it's a sample of vehicles being scrutinised, which is the population that matters most for buying decisions.
• The mileage-anomaly rate is a floor. It only counts drops greater than 1,000 miles that appear in the MOT record. Rollbacks between tests, smaller edits, or cars too new for MOT history are not captured, and some flagged cases are legitimate cluster replacements.
• Sample size: 3,626 vehicles is a solid sample for headline figures and the published brand cuts, but smaller brands fall below our thresholds and are excluded from the rankings.
• Time window: data is from August 2024 to August 2026. Manufacturer systems and dealer practices evolve; year-on-year comparisons would need a longer dataset.
• Figures are revised as the sample grows. This is a rolling two-year window, so numbers move between updates. The headline was 41% at 2,134 lookups in July 2026, 37.5% at 2,966 in early August, and is 36.5% at 3,626 today, partly from genuine sample growth and partly from the July provider-outage correction described in the methodology. We update in place rather than republishing, so always cite the figure and the sample size together.

Use this data

All findings in this report are free to cite, quote, or republish with attribution to FindServiceHistory and a link back to this page (findservicehistory.com/blog/uk-service-history-coverage-research-2026). Suggested attribution line:

Source: FindServiceHistory UK Service History Coverage Report 2026, based on 3,626 anonymised vehicle lookups (August 2024 – August 2026).

Journalists or researchers wanting the full 31-brand MOT table, additional cuts, or a custom angle can email research@findservicehistory.com. We can usually turn around a custom data pull within 48 hours.