How Long Do UK Cars Actually Last? Survival Rates by Model
Key findings
- The median UK car reaches 16.5 years old. That is the age at which a registration cohort falls below half still being MOT-tested, read across cohorts as they stand at the snapshot date of 2025-12-30. At 10 years 88.8% are still on the road; by 20 it is 19.6%, and by 22 only 10.6%.
- The spread between brands is 35.6 percentage points. Of 2010-registered cars, 82.2% of Hondas are still tested against 46.6% of Chevrolets — the difference between more than four in five surviving and fewer than one in two.
- Small cars outlast big ones, decisively. Six of the top fifteen are superminis or city cars and not one appears in the bottom fifteen, which is executive and large family saloons, seven-seat MPVs, vans and a pickup — vehicles that become uneconomic to repair while still comparatively young.
- The Vauxhall Insignia has the lowest survival rate we measured at 28.5% — fewer than three in ten of the 34,979 registered in 2010 are still being tested, against 89.0% for the Honda Jazz.
How we worked this out
This is a survival analysis, not an opinion survey. The source is DfT and DVSA’s Anonymised MOT tests and results dataset — the bulk open publication on data.gov.uk, which contains public sector information licensed under the Open Government Licence and carries no registration mark and no VIN. We read 725,145,134 class-4 test rows covering 2005–2025, with a newest test date of 2025-12-30, and grouped the 69,154,885 distinct vehicles in that series by year of first registration.
A vehicle counts as still on the road if its most recent MOT test falls on or after 2024-11-29. An MOT certificate lasts twelve months, so any car legally in use has been tested inside that window; we allow thirteen months of slack for late tests. Vehicles with no MOT test on record at all are excluded, because they never entered the MOT system and cannot be scored either way.
The survival rate for a cohort is simply vehicles still tested ÷ vehicles in the cohort. The whole-fleet curve covers 47,780,229 vehicles registered 2003–2022. Cohorts before 2003 are excluded rather than published with a caveat: the open MOT series begins in 2005 and is only fully covered from 2006, so an earlier cohort’s denominator would be not the whole cohort but the part of it that survived long enough to be recorded. The brand and model tables use the 2010 cohort — cars turning 15, old enough for real attrition to have happened but recent enough that the models are still familiar. We publish a model only where at least 3,000 of that model were registered in 2010, and a brand only above 10,000, so no ranking rests on a thin sample. That leaves 135 models and 25 brands.
The median service life of 16.5 years is the age at which the curve crosses 50%, linearly interpolated between the two adjacent cohort years — here, the 2009 cohort at 57.4% and the 2008 cohort at 43.4%. Every figure on this page is produced by scripts/research/01-fleet-survival.mjs in our repository and rendered straight from that script’s output — nothing here is typed by hand.
What this figure is, exactly. It is a cross-section: each cohort is looked at once, against the snapshot, and a car’s age is its calendar age at that moment, whenever it actually stopped being tested. That is not the only defensible way to measure a lifespan, and it is not the one we use on a vehicle report. The Estimated Lifespan panel there asks a narrower question — the median age at last MOT for cars first used 2003 or later — and answers 15.8 years, 0.7 years below this page. The two also now run off different snapshots: the report card is built from our DVSA MOT History records to 2026-04-12, this page from the open dataset to 2025-12-30. Same definition of “on the road”, different estimator, different population, different source. We reconcile the two line by line in section 4.1 of our methodology rather than quietly moving one to match the other.
What this data cannot tell you
- Survival is not reliability. It measures whether an owner keeps paying to keep the car legal, which blends fault rate, repair cost and what the car is worth. A £900 gearbox on a £1,200 car ends the car; the same bill on a £9,000 car does not.
- Left-truncated cohorts are excluded, not caveated. The open MOT series begins in 2005 and is only fully covered from 2006. A car registered before 2003 and scrapped before its first recorded test never enters the data, so its cohort’s denominator would be the part that survived to be recorded rather than the whole cohort. Earlier versions of this page published those cohorts with a warning attached; they are now simply not published, and the curve starts at 2003. No figure on this page depends on them.
- A cohort excludes cars that never had an MOT at all. A vehicle written off or exported inside its first three years never enters the MOT system and cannot be scored either way, so it is absent from both the numerator and the denominator. That nudges every cohort’s survival rate up by the same kind of margin, and it is the price of using the MOT record as the frame.
- Exports and SORN read as death. A car exported to Ireland or declared off-road leaves the MOT record and counts as not surviving even though it may still be running. This depresses vans and fleet models in particular, which are exported at high rates.
- It is a snapshot, not a live feed. The newest test in the data is 2025-12-30. Anything that happened after that date is invisible here.
- Model names are as DVSA records them. Testers enter the model from the V5C, so some appear as short codes — Mercedes-Benz C-Class as “C”, for example — and a handful of models are split across spellings, which can shrink a cohort below our publishing floor.
- We deliberately do not publish a pass or failure rate from this source. These records hold each vehicle’s latest test only, so a car that failed and then passed a retest appears as a pass. Any “failure rate” computed from it would be wrong, so we do not compute one.
Correction, 28 August 2026 — the median has moved from 17.3 to 16.5 years
Between 10 and 27 August 2026 this page reported a median UK car life of 17.3 years from 19.4 million vehicles. Both figures were wrong. The data load behind them was truncated: only 29.62% of the intended registrations were present, so the cohorts were both smaller and unrepresentatively composed.
Everything on this page has been re-derived from DVSA and DfT’s open Anonymised MOT tests and results dataset — 725,145,134 class-4 test rows covering 2005–2025. The median is now 16.5 years, the cohort range is 2003–2022 rather than 1995–2020, and the brand and model tables have changed hands: the best-surviving brand of 2010 is Honda at 82.2%, not Mini at 75.4%, and the worst is Chevrolet at 46.6%, not Vauxhall at 48.7%. The top-to-bottom spread is 35.6 points, not 26.7.
Cohorts registered before 2003 are no longer published at all. The earlier version showed them with a warning that their survival rates were upper bounds; excluding them is the honest treatment, so the curve now starts at 2003 instead of 1995. Separately, the Estimated Lifespan panel on our free vehicle reports answers a narrower question — the median age at last MOT for cars first used 2003 or later — and puts the fleet at 15.8 years, 0.7 years below this page. Full reconciliation in our methodology.
Britain scraps about a million cars a year, and almost every published answer to “how long does a car last?” is either a manufacturer’s claim or an owner-survey opinion. The MOT record answers it directly. A car that is still being tested is still on the road; a car that stopped being tested has left. Follow a whole year of registrations forward and you get a survival curve for the British car fleet.
We did that for 47,780,229 vehicles. The answer: the median UK car reaches 16.5 years old — the age at which a cohort falls below half still being tested — and which badge is on the bonnet changes that by a wide margin.
The survival curve of the British car fleet
Attrition is not steady. Almost nothing is lost in the first six years — 95.4% of the 2019 cohort was still being tested. The curve then bends sharply through the teens: 88.8% survive to ten, 64.7% to fifteen, 19.6% to twenty and 10.6% to 22. Past twenty the remaining cars are a different population altogether: classics, cherished low-mileage examples and specialist vehicles.
| Age | Registered | Cohort | Still tested | Survival |
|---|---|---|---|---|
| 3 | 2022 | 1,660,950 | 1,658,266 | 99.8% |
| 4 | 2021 | 1,800,506 | 1,750,770 | 97.2% |
| 5 | 2020 | 1,763,880 | 1,692,086 | 95.9% |
| 6 | 2019 | 2,481,308 | 2,366,690 | 95.4% |
| 7 | 2018 | 2,520,445 | 2,377,274 | 94.3% |
| 8 | 2017 | 2,734,486 | 2,490,917 | 91.1% |
| 9 | 2016 | 2,860,187 | 2,600,200 | 90.9% |
| 10 | 2015 | 2,815,335 | 2,501,510 | 88.8% |
| 11 | 2014 | 2,639,908 | 2,269,702 | 86.0% |
| 12 | 2013 | 2,401,769 | 1,979,284 | 82.4% |
| 13 | 2012 | 2,154,637 | 1,676,749 | 77.8% |
| 14 | 2011 | 2,067,252 | 1,475,616 | 71.4% |
| 15 | 2010 | 2,130,490 | 1,378,802 | 64.7% |
| 16 | 2009 | 2,079,287 | 1,192,646 | 57.4% |
| 17 | 2008 | 2,284,845 | 992,262 | 43.4% |
| 18 | 2007 | 2,611,237 | 906,237 | 34.7% |
| 19 | 2006 | 2,546,639 | 683,408 | 26.8% |
| 20 | 2005 | 2,635,212 | 515,977 | 19.6% |
| 21 | 2004 | 2,780,400 | 394,625 | 14.2% |
| 22 | 2003 | 2,811,456 | 297,647 | 10.6% |
The cars that last: survival by model
These are the 135 models with at least 3,000 UK registrations in 2010, ranked by how many are still being tested 15 years on. The pattern is hard to miss: cheap-to-run small cars at the top, expensive-to-run large ones at the bottom.
| # | Model | Registered 2010 | Still tested | Survival |
|---|---|---|---|---|
| 1 | Mazda Mx-5 | 4,876 | 4,458 | 91.4% |
| 2 | Honda Jazz | 23,809 | 21,191 | 89.0% |
| 3 | Volkswagen Transporter | 11,861 | 10,534 | 88.8% |
| 4 | Audi Tt | 7,150 | 6,142 | 85.9% |
| 5 | Land Rover Freelander | 15,404 | 13,175 | 85.5% |
| 6 | Nissan Juke | 5,251 | 4,400 | 83.8% |
| 7 | Toyota Yaris | 21,249 | 17,790 | 83.7% |
| 8 | Mercedes-Benz A-Class | 6,811 | 5,682 | 83.4% |
| 9 | Toyota Iq | 4,379 | 3,636 | 83.0% |
| 10 | Fiat Ducato | 3,911 | 3,230 | 82.6% |
| 11 | Honda Cr-V | 11,404 | 9,394 | 82.4% |
| 12 | Toyota Aygo | 13,301 | 10,950 | 82.3% |
| 13 | Mercedes-Benz B-Class | 5,109 | 4,175 | 81.7% |
| 14 | Nissan Note | 16,214 | 13,232 | 81.6% |
| 15 | Fiat 500 | 22,647 | 18,363 | 81.1% |
Not everything at the top is small, and the exceptions are instructive: the Mazda Mx-5 that leads the table is a two-seat sports car kept for weekends rather than commuting, the Volkswagen Transporter is a van, and the Land Rover Freelander and Honda CR-V are mid-size 4×4s — all vehicles that stay worth repairing long after a saloon of the same age is not. Among ordinary family cars the Honda Jazz is the standout: 89.0% of the 23,809 registered in 2010 are still being tested. Toyota places 3 models in the top fifteen. The city-car triplets sold as the Toyota Aygo, Peugeot 107 and Citroen C1 — the same car with three badges — land within 1.9 points of each other at 82.3%, 81.1% and 80.4%. That near-identical result on three different badges is a useful check that the method is measuring the car rather than the brand’s reputation.
The cars that disappear first
| # | Model | Registered 2010 | Still tested | Survival |
|---|---|---|---|---|
| 1 | Vauxhall Insignia | 34,979 | 9,952 | 28.5% |
| 2 | Mazda 6 | 9,960 | 4,121 | 41.4% |
| 3 | Peugeot 5008 | 3,755 | 1,589 | 42.3% |
| 4 | Renault Grand Scenic | 7,745 | 3,360 | 43.4% |
| 5 | Citroen C4 | 10,763 | 4,778 | 44.4% |
| 6 | Vauxhall Zafira | 29,314 | 13,383 | 45.7% |
| 7 | Toyota Hilux | 4,624 | 2,116 | 45.8% |
| 8 | Peugeot Bipper | 5,547 | 2,553 | 46.0% |
| 9 | Ford Mondeo | 28,294 | 13,032 | 46.1% |
| 10 | BMW 520 | 7,571 | 3,593 | 47.5% |
| 11 | Citroen Dispatch | 3,612 | 1,753 | 48.5% |
| 12 | Peugeot 308 | 20,560 | 9,983 | 48.6% |
| 13 | Ford Transit | 26,029 | 12,824 | 49.3% |
| 14 | Skoda Superb | 3,092 | 1,561 | 50.5% |
| 15 | Skoda Octavia | 15,861 | 8,047 | 50.7% |
The bottom fifteen is executive and large family saloons, seven-seat MPVs, vans and a pickup. There is not a single supermini in it. The Vauxhall Insignia sits last at 28.5%: of 34,979 registered in 2010, 9,952 are still tested. Sold overwhelmingly as company cars, they covered enormous mileages early and reached the point where a repair exceeded the car’s value while still relatively young.
Model averages can't tell you about one specific car
See the full MOT history of the car you're looking at
Every test, every advisory and every mileage reading DVSA holds for that registration — the record that actually shows whether a car has been looked after.
- Every pass, fail and advisory since 2005
- Mileage at each test, with gaps and rollbacks flagged
- Outstanding safety recalls
Free. Official DVSA data. No sign-up.
Survival by brand
Rolling the same cohort up to brand level removes most of the model-mix noise and leaves a 35.6-point spread from top to bottom.
| # | Brand | Registered 2010 | Still tested | Survival |
|---|---|---|---|---|
| 1 | Honda | 63,702 | 52,336 | 82.2% |
| 2 | Mini | 42,122 | 32,307 | 76.7% |
| 3 | Nissan | 93,943 | 70,823 | 75.4% |
| 4 | Toyota | 93,338 | 69,760 | 74.7% |
| 5 | Land Rover | 41,146 | 30,550 | 74.2% |
| 6 | Suzuki | 20,622 | 15,091 | 73.2% |
| 7 | Hyundai | 58,495 | 41,568 | 71.1% |
| 8 | Volkswagen | 191,176 | 135,403 | 70.8% |
| 9 | Fiat | 60,489 | 42,438 | 70.2% |
| 10 | Volvo | 36,741 | 25,535 | 69.5% |
| 11 | Audi | 93,012 | 63,879 | 68.7% |
| 12 | Ford | 309,432 | 207,061 | 66.9% |
| 13 | Jaguar | 16,000 | 10,698 | 66.9% |
| 14 | Kia | 53,936 | 35,846 | 66.5% |
| 15 | Mazda | 44,689 | 29,408 | 65.8% |
| 16 | Mercedes-Benz | 75,113 | 47,581 | 63.3% |
| 17 | Skoda | 39,453 | 24,250 | 61.5% |
| 18 | BMW | 104,158 | 63,323 | 60.8% |
| 19 | Citroen | 85,907 | 51,122 | 59.5% |
| 20 | Renault | 103,973 | 60,165 | 57.9% |
| 21 | Mitsubishi | 18,414 | 10,540 | 57.2% |
| 22 | Peugeot | 119,143 | 66,574 | 55.9% |
| 23 | Seat | 30,290 | 16,805 | 55.5% |
| 24 | Vauxhall | 264,544 | 134,391 | 50.8% |
| 25 | Chevrolet | 13,381 | 6,237 | 46.6% |
Two results deserve a caveat rather than a headline. Land Rover scores 74.2% despite a poor reputation in owner surveys — a Defender or Freelander is worth repairing at an age when a hatchback is not, and survival rewards that. Chevrolet’s last place rests on a thin and unusual base and should be read with care: of its 13,381 cars registered in 2010, only the Spark clears our 3,000-car model floor, and that one survives at 59.3% — well above the brand figure. The rest sits in models too small to publish. Second from bottom, Vauxhall at 50.8% is the clearer mix effect: its 2010 volume was concentrated in the Astra, Zafira, Insignia and Combo, all of which sit in the bottom half of the model table on their own.
The consistent, mix-independent finding is the one visible in the model table: size and replacement cost predict survival better than brand reputation does. If you want a car that is still worth keeping at 15, the strongest signal in this data is to buy a small one.
What this means if you are buying used
A model’s survival rate tells you the odds; it does not tell you about the car in front of you. Two 2010 Fiestas with identical survival odds can be in completely different condition, and the MOT record shows which is which. What to look for:
- Unbroken annual testing. A gap of eighteen months or more usually means the car sat unused or was uneconomic to fix at the time.
- Repeat advisories on the same item. A corrosion advisory that reappears year after year is the single clearest predictor of the repair that eventually ends the car — see our guide to rust and MOT failure.
- Mileage that moves normally. Compare the readings against the real mileage distribution for that age. Very low mileage on an old car is often good; a mileage that falls between tests is a clocking warning.
Half of a UK registration cohort is gone by 16.5
Find out where the car you're looking at sits
Enter the registration to see its full DVSA test history, advisory pattern and mileage record — the evidence behind whether this particular car is a survivor.
- Full MOT history, free
- Advisory patterns that predict the next big bill
- Mileage anomalies flagged automatically
Free. Official DVSA data. No sign-up.
Cite this research
harvardUse this citation when referencing this report in articles, papers, or blog posts. Refreshes the access date each time the page loads.
MOT Checkup (2026) 'How Long Do UK Cars Actually Last? Survival Rates by Model'. Available at: https://www.motcheckup.co.uk/research/how-long-uk-cars-last-2026 (Accessed: 10 September 2026).
Download the data
The full How Long Do UK Cars Actually Last? Survival Rates by Model dataset is available free. MOT Checkup's compilation and analysis are released under CC BY 4.0 — reuse with attribution to MOT Checkup. That grant covers our aggregation and presentation, not the underlying source records.
Using this research
Journalists, researchers and site owners are welcome to reproduce any table or chart on this page with attribution to MOT Checkup. Our analysis, tables and charts are released under CC BY 4.0. They are computed by MOT Checkup from the DVSA/DfT “Anonymised MOT tests and results” dataset published on data.gov.uk, which contains public sector information licensed under the Open Government Licence v3.0. The CC BY grant covers our aggregation and presentation; the underlying records remain under the OGL. If you need a cut of the data we have not published — a specific model, a different age band, a regional split — email hello@motcheckup.co.uk and we will run it.
Frequently asked questions
How long does the average UK car last?
The median British car reaches 16.5 years old. Precisely: that is the age at which a first-registration cohort falls below half still being MOT-tested, measured across 47,780,229 vehicles registered 2003–2022 and checked against the 13 months to 2025-12-30. The crossing is interpolated between the 2009 cohort (57.4% still tested at 16) and the 2008 cohort (43.4% at 17).
Why does your vehicle report say a car's model is half gone by around 15.8 years?
Because the report card measures a different thing: the median AGE AT LAST MOT for cars first used 2003 or later, built as an actuarial life table in which a retirement is dated at the age of that car's own final test and still-tested cars are right-censored. On that basis the fleet crosses 50% at 15.8 years. This page measures a cross-section instead: every cohort looked at once against a single snapshot, with a car's age taken as its calendar age at that moment. The two answer different questions and sit 0.7 years apart. Neither is a correction of the other — the full reconciliation is on our methodology page.
Which car lasts the longest in the UK?
Of the 135 models with enough 2010-registered cars to measure reliably, the Mazda Mx-5 has the highest survival rate at 91.4% — 91.4% of the ones registered in 2010 were still being MOT-tested 15 years later. Among ordinary family cars the Honda Jazz leads at 89.0%.
Which car brand lasts longest in the UK?
Honda tops the brand table for the 2010 cohort at 82.2%, ahead of Mini (76.7%) and Nissan (75.4%). Chevrolet is last at 46.6%. The gap between the top and bottom brand is 35.6 percentage points.
Does a high survival rate mean a car is reliable?
Not on its own. Survival measures whether owners keep paying to keep a car legal — which is reliability, repair cost and residual value combined. A cheap car with a £900 repair bill gets scrapped; an expensive car with the same bill gets fixed. That is why big executive cars and vans sit near the bottom of this table despite not being unusually fault-prone. Read it as a running-cost signal, not a fault rate.
Why do vans and fleet cars score so badly?
Vehicles bought as fleet or business transport — Vauxhall Insignia, Ford Mondeo, Ford Transit, BMW 520, Skoda Octavia — cover very high mileages in their first few years and are commonly exported or broken for parts rather than sold on again. They leave the UK MOT record early. A vehicle exported while still perfectly serviceable reads as "not surviving" in this dataset.
How can I tell whether a specific used car is a survivor?
Model averages cannot tell you about the individual car in front of you. The MOT record can: it shows every test the vehicle has had, every advisory the tester logged, and the mileage at each visit. A car with continuous annual tests and few repeated advisories has been kept up; long gaps and recurring corrosion advisories are the pattern that precedes a scrapping decision. Run the registration through a free MOT history check before you buy.
How often is this research updated?
We refresh the analysis when a new MOT records snapshot is loaded. The next scheduled review is February 2027, which will extend every curve by one year and bring one more registration cohort into the uncensored range.
Sources and further reading
- 1.DVSA/DfT — Anonymised MOT tests and results (the open bulk dataset this analysis is computed from) — Primary data (Open Government Licence)
- 2.Open Government Licence v3.0 — Licence
- 3.DVSA — MOT testing data for Great Britain — Official statistics
- 4.Department for Transport — vehicle licensing statistics (VEH0 tables) — Official statistics
- 5.DVSA MOT inspection manual — cars and passenger vehicles — Reference
- 6.MOT Checkup — free MOT history check by registration — Tool
- 7.MOT Checkup — mileage and clocking check — Tool
- 8.MOT Checkup — common MOT faults by make and model — Tool
About this research
MOT Checkup publishes original analysis of UK vehicle data. Every figure on this page was produced by a script committed to our repository and run against the source records described above — we do not publish estimates, modelled figures or numbers we cannot reproduce. Where an analysis did not hold up, we dropped it rather than soften it. Spotted an error? Tell us and we will correct the page and note the correction.
More checks before you buy
Run all of these on any UK car before you pay a deposit — every one is free to start.
Full Car Check
MOT, mileage, tax and spec by reg — the free starting point.
MOT History Check
Every test, advisory and failure on the DVSA record — free.
DVLA Vehicle Check
The DVLA registration record and DVSA MOT history in one search.
Car Import Check
No free import marker exists — how to read the four indicators instead.
Mileage / Clocking Check
Spots odometer rollbacks and implausible jumps across MOTs.
MOT Failure History
Pass-rate and recurring-failure patterns from MOT history.
Outstanding Finance
Why unpaid finance can cost you the car — and how to verify.
Stolen Car Check
The warning signs of a stolen vehicle before you pay a deposit.
Write-Off Check
Cat A/B/S/N explained and what a write-off means for buyers.
Cat S / Cat N Check
Insurance write-off categories explained — is it worth buying?
HPI Check Alternative
What each check covers and costs — an honest price comparison.
Car Tax Check
Tax status, due date and SORN by registration — free.
Insurance Group Check
The 1–50 insurance group for any model — what it will cost to cover.
Is My Car Insured?
How to confirm a car's insurance status before you drive it away.
Car Tax Calculator
Estimate the annual VED road tax for any car by reg or spec.
Car Recall Check
Check DVSA safety recalls for the make and model, free.
Selling Your Car
See the record a buyer will pull on your reg — before they pull it.
Service History: Who Holds It
No site can look one up. Who really has the record, by make.
Used Car Buying Checklist
The full step-by-step checklist before you buy a used car.
Buying a Cat S / Cat N Car
Is a write-off worth it? The honest costs, risks and decision rule.
Check Finance Before Buying
Why there's no free finance check — and how to verify before you pay.
What Does HPI Clear Mean?
Why an 'HPI clear' result does NOT mean accident-free.
How to Spot a Clocked Car
The tell-tale signs of mileage fraud before you buy.
What Is a Cat S Car?
Category S write-off meaning — structural damage, explained.
What Is a Cat N Car?
Category N write-off meaning — and why 'non-structural' isn't 'minor'.