Longest Lasting EVs Study 2026: 74 Models Ranked on Longevity and Battery Health

The 2026 headline findings#

7.9% of electric cars on British roads have reached 150,000 miles and 24.6% have reached 100,000. Mean battery state of health at 100,000 miles is 90.4%, and the single largest controllable influence on it is the share of energy taken from rapid chargers, worth 10.6 percentage points.

This study has two halves and they answer different questions. The first asks which electric cars last: the share of each model that has passed 100,000, 150,000 and 200,000 miles, and the mean mileage at which a car of that model leaves the road. The second asks what makes any electric car last, which is a question about charging behaviour rather than about badges. The second half is the genuinely new material, because it is built on twelve months of logged charging data rather than on owner recollection.

One fact has to be stated before any of the rankings, because everything else depends on it. The UK electric car parc is young. The EV Cable Hub UK EV Lifecycle Panel 2026 puts the mean age of an electric car in Britain at 3.8 years, which means the observed share that has reached any high mileage understates the share that eventually will. That is why every high-mileage figure on this page is published twice: once as observed, and once age-adjusted. Observed, 7.9% of UK electric cars have passed 150,000 miles. Age-adjusted, the projection is 18.4%. Both numbers are real and they measure different things, and a study that published only one of them would deserve to be dismissed.

The model spread is wide. The Tesla Model S reaches 150,000 miles in 26.8% of cases and the Mazda MX-30 in 3.2%, a ratio of 8.4 to 1. That gap is not a durability ranking, and the section that publishes it says so plainly: it is heavily shaped by who buys each car and how far they drive it. The battery health ranking, which holds mileage constant at 100,000 miles, is the durability measure, and it produces a completely different order.

The behavioural findings are where the practical value sits. Taking more than half your energy from rapid chargers costs 10.6 percentage points of battery capacity by 100,000 miles. Charging habitually to 100% costs 5.2 points. Preconditioning before a winter rapid charge is worth 2.6 points. Home charging power, the variable most owners assume matters, moves battery health by 3.0 points across every power from 2.3kW to 22kW, and by only 1.4 points once annual mileage is controlled for. This study sits alongside three others built from the same dataset: the Fastest Selling EVs Study 2026, the EV Depreciation Study 2026 and the Used EV Price Index 2026, all four gathered under the UK EV Lifecycle Panel hub.

One convention runs through every number on this page and it is worth stating once. Longevity here is measured in two units that are not interchangeable: the share of a model that has passed a mileage threshold, and the battery capacity a model still holds at a fixed mileage. The first is dominated by who buys the car and how far they drive it. The second is a property of the car. Almost every disagreement about which electric cars last longest comes down to which of those two things the speakers are measuring, and this study publishes both, in full, for all 74 models.

The full ranking, all 74 models by share reaching 150,000 miles#

The Tesla Model S is the longest lasting electric car in Britain, with 26.8% of examples reaching 150,000 miles in 2026. The Mazda MX-30 is the shortest lived at 3.2%, and the gap between them is a factor of 8.4.

Read the table this way. The share reaching each mileage band is the proportion of surveyed and resale-captured examples of that model that have passed the threshold. The basis column states whether the model has a cohort old enough for that figure to be observed rather than projected: 50 of the 74 models carry observed figures and 24 carry projected ones. A projected figure is a model output, not a measurement. It applies that model's observed mileage accumulation rate and observed exit hazard to the current age distribution of its UK parc, and it should be read with the same caution as any forecast.

The pattern at the top of the table is consistent and it has two causes rather than one. Early, long-range models were bought disproportionately by high-mileage drivers, so they accumulate distance faster than the parc average. And a large battery covers the same annual mileage in fewer full-equivalent cycles than a small one, so the same journey is gentler on a 95kWh pack than on a 30kWh pack. The EV Cable Hub UK EV Lifecycle Panel 2026 found both effects present in the top ten, which is why the first four rows are Teslas and why the fifth and sixth are two of the oldest small cars in the British parc rather than two of the most durable ones.

The bottom of the table needs stating carefully, because the obvious reading of it is wrong. The lowest rows are dominated by small-battery city cars bought by low-mileage drivers. A Mazda MX-30 rarely reaches 150,000 miles because a Mazda MX-30 is rarely driven 150,000 miles, not because it wears out at 149,000. That is a usage finding, not a durability finding, and it is the criticism a careful reader would otherwise make of this table. The durability measure is battery state of health at a fixed mileage, and on that measure the MX-30 sits 67th of 74 rather than last, a materially different position from the one it occupies here.

One further caveat belongs with this table rather than in the methodology. Vehicles used at some point for private hire or taxi work are 4.6% of the Panel but 74.3% of the group above 250,000 miles, and they are concentrated in a small number of large, early models. They are included, because excluding them would understate what those cars actually achieve, but their influence on the top ten rows is real and it is stated here rather than buried.

The mileage bands, 100,000 to 250,000#

24.6% of UK electric cars have reached 100,000 miles in 2026, 7.9% have reached 150,000, 2.1% have reached 200,000 and 0.6% have reached 250,000. Age-adjusted, the projected shares are 46.2%, 18.4%, 6.8% and 2.4%.

The age adjustment is a single, stated calculation rather than a judgement. It applies each model's observed mileage accumulation rate and its observed exit hazard to the current age distribution of that model's UK parc, and reports the share that would pass each threshold if today's fleet aged on today's behaviour. It is a projection and it is labelled as one everywhere it appears. The gap between the two columns is largest in the middle bands, because that is where the young parc bites hardest: 7.9% of cars have passed 150,000 miles today, but 18.4% are projected to.

Each threshold means something different to a buyer. 100,000 miles is the point at which most battery warranties expire on mileage rather than on years, and 9,573 vehicles in the Panel have passed it at a mean age of 5.8 years and a mean battery state of health of 90.4%. 150,000 miles is where a petrol car is usually considered to be near the end of an economic life; 3,074 Panel vehicles have passed it, at a mean age of 7.8 years and 86.1% health. 200,000 miles is genuinely rare at 817 vehicles, and the mean battery there still holds 82.4% of its original capacity.

The 250,000-mile group is small enough to describe individually. The EV Cable Hub UK EV Lifecycle Panel 2026 contains 148 vehicles above 250,000 miles, at a mean age of 10.8 years and a mean battery state of health of 79.2%. They are not a random sample of the parc: 61.5% are a Tesla Model S or Model X, 12.8% are a Nissan Leaf and 6.1% are a BMW i3, and 74.3% have been used for private hire or taxi work at some point. Their mean annual mileage is 24,860, roughly three times the parc average. 22.3% have had a battery replaced and 18.9% a drive unit. The highest odometer reading recorded anywhere in the Panel is 418,640 miles, on a 2015 Tesla Model S that has had two battery replacements and still reads 71.4% state of health.

The band table also answers a question the model rankings cannot, which is what condition a high-mileage electric car is actually in when you find one. Mean battery state of health falls from 94.2% at 50,000 miles to 90.4% at 100,000, 86.1% at 150,000 and 82.4% at 200,000, and mean age at each band rises from 3.4 to 9.4 years. A 150,000-mile electric car in Britain is, on the Panel average, a seven-to-eight year old car holding around 86% of its original range. That is a far more useful picture for a used buyer than any single survival percentage.

Mean lifetime mileage, all 74 models ranked#

The average electric car leaves the UK parc at 78,400 miles in 2026. The Tesla Model S goes furthest at 102,400 miles and the Mazda MX-30 the least far at 54,200, a difference of 48,200 miles.

Lifetime mileage is defined precisely here because the phrase is used loosely elsewhere. It is the mean odometer reading at the point a vehicle leaves the UK parc through scrappage, insurance write-off or export, taken from the exit records in the EV Cable Hub Listing Tracker 2026 for models with sufficient exit data, and projected from the mileage accumulation curve for the rest. It is not a measure of how far a car could go. It is a measure of how far the cars of that model actually went.

That makes it a measure of use as much as of engineering, and both matter to a buyer. A car that is bought by people who drive 15,000 miles a year will show a higher lifetime mileage than an identically durable car bought by people who drive 5,000, and no amount of statistical care removes that. The value of the measure is that it answers the question an owner actually has, which is how much life is realistic for the car in front of them given how cars like it get used.

The exit table carries the most reassuring finding in this study and it should be stated flatly rather than sold. The most common reason an electric car leaves the British parc is an insurance write-off after a collision, at 46.8% of exits, at a mean of 62,400 miles and 4.9 years with the battery still at 93.2% health. Export accounts for a further 21.4%. Battery-related exit, meaning an uneconomic battery repair or replacement, accounts for 8.2%, at a mean of 126,400 miles and 9.1 years. The EV Cable Hub UK EV Lifecycle Panel 2026 found that the failure mode the public associates with electric cars is the fourth most common way one leaves the road, well behind a crash.

Battery state of health by mileage#

A UK electric car holds 90.4% of its original battery capacity at 100,000 miles in 2026 and 86.1% at 150,000. The steepest decline happens in the first 20,000 miles, where cars lose 3.1 percentage points, after which the curve settles to roughly 1.9 points per 25,000 miles.

The early-life step is the part of the curve that surprises owners, and it has a straightforward explanation. A new pack loses a measurable slice of capacity in its first months as the cell chemistry settles, and that loss is front-loaded rather than spread evenly. The Panel curve shows 1.1 points gone by 5,000 miles and 3.1 points by 20,000. An owner who checks state of health at 18 months and finds 96% has not been sold a faulty car; they have watched the normal shape of the curve.

After that the decline is close to linear and considerably gentler than the early step implies. Between 20,000 and 100,000 miles a car loses a further 6.5 percentage points, which is less than a fifth of a point per 5,000 miles. The EV Cable Hub UK EV Lifecycle Panel 2026 recorded a mean of 86.1% at 150,000 miles, 82.4% at 200,000 and 79.2% at 250,000. Anyone extrapolating from the first year of ownership overestimates lifetime degradation by a wide margin, which is the most common error made about this data.

The number a sceptical reader wants is the share of cars below the level at which a battery warranty pays out, and it should be stated rather than left to inference. Most published warranty thresholds sit at 70% of original capacity. The mean Panel curve does not reach 70% anywhere inside the observation window, and 0.9% of all Panel vehicles sit below it, falling to 0.4% inside eight years. Those vehicles have a mean age of 9.8 years and a mean mileage of 168,400. The share below 80% is more useful as a practical measure and it rises with mileage in a predictable way: 1.2% at 50,000 miles, 4.4% at 100,000, 11.4% at 150,000 and 34.8% at 250,000.

Spread matters as much as the mean, because a mean conceals the car a buyer might actually be looking at. At 100,000 miles the best decile of the Panel holds 93.8% and the gap between the best and worst decile is 8.8 percentage points. Of that spread, 61.4% is explained by charging pattern, 18.2% by chemistry, 6.8% by climate and region, and 13.6% is unexplained by the variables tested. That single breakdown is the argument of this entire study: how a car is charged matters roughly three times as much as what chemistry is inside it, and more than nine times as much as where in Britain it lives.

It is worth being precise about what a state-of-health reading is and is not. It is a measure of usable capacity against the capacity the pack had when new, not a measure of whether the car works. A pack at 86% drives exactly as it did at 100%, accelerates the same and charges at the same power; it simply covers 14% less distance between charges. That distinction matters because degradation is routinely described in language borrowed from mechanical failure, and the two behave nothing like each other. A worn engine gets worse to drive. A worn battery gets shorter.

Battery health by model, all 74 ranked#

The BYD Dolphin holds the most battery capacity at 100,000 miles of any electric car in Britain, at 92.6% in 2026. The Nissan Leaf 40kWh holds the least at 79.6%, a spread of 13.0 percentage points.

This is the durability ranking, and it should be read instead of the mileage ranking by anyone asking which electric car has the better battery. Mileage is held constant at 100,000 miles, so how far the car is driven and who drives it drop out of the comparison entirely. The order it produces bears very little resemblance to the order in the mileage table: the Tesla Model S, first for reaching 150,000 miles, sits 32nd here, and the Nissan Leaf 40kWh, fifth for reaching 150,000 miles, sits last.

Two mechanisms drive the ranking and both are visible in the table's own columns. The first is chemistry. Seven of the top ten models run lithium iron phosphate, which degrades more slowly than the nickel-based chemistries under the same use. The second is thermal management. All ten of the top ten use active liquid battery cooling, and both of the two lowest use passive air cooling. Six of the 74 models in the Panel are passively cooled, and five of those six sit in the bottom eight.

The thermal management comparison is the cleaner of the two because it splits the Panel into two groups of very unequal size but very consistent behaviour. The EV Cable Hub UK EV Lifecycle Panel 2026 recorded 91.2% mean health at 100,000 miles for actively liquid-cooled packs against 82.4% for passively cooled ones, a gap of 8.8 percentage points that widens to 10.2 points by 150,000 miles. The share of cars that have fallen below 80% health by 100,000 miles is 2.8% among liquid-cooled packs and 24.6% among passively cooled ones. That is close to a ninefold difference in the outcome an owner actually cares about, and it is the strongest single argument in this dataset that pack design, not brand, decides battery life.

One note on how to compare this table with the headline figure. The 90.4% Panel mean is weighted to the UK parc, so it reflects how many of each model are actually on the road. The model rows in this table are unweighted, so a model with 200 examples counts the same as one with 20,000. Averaging the rows of this table gives 87.2%, and the difference between the two numbers is composition rather than disagreement: the models that sell in volume are disproportionately the liquid-cooled ones.

Nothing in this table should be read as a verdict on a manufacturer. Thermal management and chemistry are specification decisions taken years before a car reaches a British driveway, and several of the models at the bottom of this ranking were designed when passive cooling was a normal choice at their price. The finding is about pack architecture, and the clearest evidence for that is that the same manufacturer appears near both ends of the table depending on which of its models is being counted.

Battery chemistry and how long a battery lasts#

LFP batteries held 91.3% of their capacity at 100,000 miles in 2026, against 86.5% for NMC and 87.2% for NCA. LFP cars were also the least sensitive to rapid charging, losing 6.2 percentage points from heavy DC use against 12.8 points for NMC.

Three chemistries cover the British electric car parc and they are worth describing without jargon. Lithium iron phosphate, written LFP, uses iron and phosphate in place of nickel and cobalt; it stores less energy for a given weight but tolerates a full charge and a hard charge far better. Nickel manganese cobalt, written NMC, is the mainstream choice and covers 60 of the 74 models in the Panel; it is denser and gives more range per kilogram, and it is more sensitive to heat and to sitting at a high state of charge. Nickel cobalt aluminium, written NCA, is a close cousin of NMC used in the two oldest Teslas in the ranking, with similar behaviour and slightly better cycle life.

The durability gap is consistent across every measure the EV Cable Hub UK EV Lifecycle Panel 2026 tested. LFP holds 4.8 percentage points more capacity than NMC at 100,000 miles and 4.8 points more at 150,000. It reaches 80% health after 3,240 full-equivalent cycles against 2,180 for NMC and 2,420 for NCA, which is roughly 50% more cycles for the same amount of wear. And it loses 6.2 points from heavy rapid charging where NMC loses 12.8, which is the difference between a chemistry that shrugs off a motorway life and one that records it.

The trade-off has to be published alongside the advantage or the finding is not honest. LFP loses more usable range in cold weather: the Panel recorded a 14.6% reduction against a 20°C baseline for LFP against 9.8% for NMC. It is also less energy dense, so an LFP car generally carries less range for the same pack weight and cost. A buyer choosing LFP is trading winter range and outright range for durability and rapid-charge tolerance, and depending on how they drive that is either an obvious trade or a bad one. The same chemistries show a 6.8 percentage point gap in three-year value retention in the EV Depreciation Study 2026, which suggests the used market has started to price durability.

Two figures from the survey put the whole section in perspective. Only 21.4% of owners could state their car's battery chemistry, and only 12.8% charge according to the guidance that applies to it. Manufacturers permit routine charging to 100% on 78.6% of LFP models and advise against it on 84.2% of NMC models, so the single most consequential piece of chemistry knowledge is also the one most owners do not have. That is a labelling failure rather than an owner failure, and it is cheap to fix.

Rapid charging and what it actually costs your battery#

Electric cars taking more than half their energy from rapid chargers held 81.2% of their battery capacity at 100,000 miles in 2026, against 91.8% for cars taking less than a tenth. That 10.6 percentage point gap is the single largest controllable influence on battery life in this study.

The relationship is monotonic across every band, which is what separates a real effect from a noisy one. Health at 100,000 miles falls from 91.8% under 10% DC, to 90.4% at 10 to 19%, 88.6% at 20 to 29%, 86.8% at 30 to 39%, 84.6% at 40 to 49% and 81.2% at 50% and above. The same ordering holds at 50,000 miles, where the spread is 6.9 points, and at 150,000 miles, where it has widened to 13.0. Nothing about the shape suggests a threshold below which rapid charging is free; it suggests a dose response.

Two confounders have to be addressed openly, because a technically literate reader will raise them within a paragraph. First, cars that rapid charge heavily also cover far more miles per year: mean annual mileage runs from 6,840 in the lowest DC band to 22,860 in the highest, a difference of 16,020 miles. Second, heavy rapid charging is concentrated in a small number of high-mileage models, so part of what looks like a charging effect is a model effect. Controlling for annual mileage cuts the raw 10.6 point gap to 8.8 points. Controlling for mileage and model together cuts it to 7.4 points. That is 69.8% of the raw effect surviving both controls, and the same pattern holds for the narrower comparisons in the table.

Publishing the controlled figure next to the raw one is the point of the section. The honest headline is not that rapid charging costs 10.6 points, it is that the difference between a car that rapid charges rarely and one that rapid charges constantly is 10.6 points, of which about seven and a half are attributable to the charging itself and the rest to the kind of car and the kind of mileage that go with it. Both numbers are useful. Only one of them is a causal claim.

Session count tells the same story through a different lens and adds the mechanism. The EV Cable Hub Charging Behaviour Panel 2026 logged health at 100,000 miles of 91.6% for cars taking fewer than six DC sessions a year and 80.8% for cars taking 96 or more. Mean pack temperature at the end of a session rises in step, from 31.4°C to 44.6°C, alongside mean session peak power rising from 68kW to 124kW. Sustained pack temperature is the mechanism, not the DC connector, and that is why the temperature exposure data later on connects back to this section. The practical counterweight is that 84.6% of the average owner's energy is already taken at home on AC, which is both the part they control most easily and the part that does the least damage. Our companion study on the impact of fast charging on battery health takes the session-level data apart in more detail.

There is a practical reading of this section that stops short of telling anyone to avoid rapid charging, and it is the one the data supports. Rapid charging is what makes an electric car usable on a long journey, and the cars in the heaviest DC band are covering 22,860 miles a year, which is work that has to be done somehow. The finding is not that rapid charging is bad; it is that rapid charging as a routine substitute for home charging is expensive in capacity terms, and that the cost is now quantified. An owner who rapid charges on trips and charges at home the rest of the time sits in the under-10% band, which is where the best battery outcomes in this dataset are.

Home charging power, cables and battery life#

Home charging power made almost no difference to battery life in 2026. Cars habitually charged at 2.3kW on a granny charger held 92.4% of capacity at 100,000 miles and cars charged at 22kW held 89.4%, a spread of just 3.0 percentage points against 10.6 points for rapid charging.

The finding needs stating flatly and then leaving alone. Within the range of power a British home can supply, charging power is close to irrelevant to battery life. A 7.4kW cable does not wear a battery meaningfully faster than a 2.3kW granny charger. The whole spread from the slowest domestic socket to a 22kW three-phase supply is 3.0 percentage points at 100,000 miles, against 10.6 points for the rapid charging variable in the previous section.

The second finding is the one that has not been published elsewhere, and it shrinks the first one further. Owners who install high-power home charging also drive further: mean annual mileage runs from 5,240 at 2.3kW to 13,860 at 22kW. Control for annual mileage and the 3.0 point spread falls to 1.4 points. A residual effect of 1.4 percentage points at 100,000 miles is smaller than the measurement uncertainty on a single state-of-health reading, and the EV Cable Hub UK EV Lifecycle Panel 2026 reports it as such rather than dressing it up. If you are choosing home charging equipment, choose it on convenience, cost and how much energy you need to move overnight, not on battery health.

The habits that do matter are all about where the battery sits rather than how fast it fills. Charging to 80% or less habitually is worth 92.4% health at 100,000 miles against 87.2% for owners who charge to 100%, a 5.2 point difference and the largest single habit effect in the table. Rarely going below 20% state of charge is worth 91.6% against 88.4% for owners regularly running below 10%. Storing a car at 50 to 60% when it is unused for more than a week is worth 92.8%, the best figure of any habit measured, though only 8.6% of owners do it. Leaving the car plugged in when parked at home is worth 91.2% against 89.8% for plugging in only when charge is needed, which inverts a widely held belief.

The cable finding is the genuinely counterintuitive one. Owners whose home charging cable reaches comfortably with slack plug in 3.8 times a week and hold 91.4% health at 100,000 miles. Owners whose cable is too short, so the car has to be repositioned to charge, plug in 2.1 times a week and hold 88.6%. That is 1.8 times the plug-in frequency at a materially better battery outcome, and it points the same way as the plugged-in-when-parked finding: more frequent low-power charging is associated with better battery health, not worse, because it keeps the pack away from both ends of its range. The mechanism is convenience. A cable that does not quite reach does not get used, and a car that does not get plugged in sits at whatever state of charge it happens to be at. Our guide to charging cable amps covers how to match a cable to a supply, and our ranges of EV charging cables, granny chargers and V2L adapters cover the equipment itself.

The equipment implication is small and should be stated as small. Nothing in this dataset supports buying a slower cable to protect a battery, and nothing supports paying to avoid a faster one. What the cable data does support is buying one long enough for where the car actually parks, because a cable that does not comfortably reach measurably reduces how often the car gets plugged in, and that shows up in battery health through the state of charge the car spends its life at rather than through the power it charges at. That is an unusual mechanism and it is the reason this finding has not appeared elsewhere.

Climate, temperature and region#

The regional spread in UK battery health is 1.7 percentage points at 100,000 miles in 2026. Northern Ireland recorded the highest mean at 90.9% and Greater London the lowest at 89.2%, and the effect of UK climate is small next to the 10.6 points attributable to charging pattern.

The honest framing of this section is the finding itself: within the United Kingdom, climate is a minor variable. Every one of the twelve regions sits between 89.2% and 90.9%, and the whole spread is smaller than the effect of a single charging habit. Region explains 6.8% of the variation in battery health between cars, against 61.4% explained by charging pattern. A buyer choosing between an identical car in Inverness and one in Southampton should not be choosing on battery health.

The regional ordering is also not really about weather. Greater London sits last with a mean DC share of 13.4%, the highest of any region, and Northern Ireland sits first with 9.4%, the lowest. The regions run in almost exactly the same order on both measures. What looks like a north-south climate gradient is mostly a charging-behaviour gradient, driven by how much off-street parking each region has and how much of its charging therefore happens on the public rapid network.

The temperature exposure table carries the more interesting material and it connects directly back to rapid charging. The EV Cable Hub Charging Behaviour Panel 2026 logged a Panel mean of 62 hours a year with the pack above 35°C and 168 hours below 0°C. Hours above 35°C correlate with capacity loss at 0.64. Hours below 0°C correlate at 0.12. Cold, in other words, costs range on the day and very little capacity over the life of the pack, while sustained heat costs capacity permanently. British ambient conditions almost never push a pack above 35°C on their own. Rapid charging does, which is why the high-rapid rows in the exposure table sit 5.7 to 6.6 points below their low-rapid equivalents whatever the parking situation, and why the whole exposure spread is 8.9 points.

Parking situation matters within that, though less than the exposure table's headline spread suggests. Comparing like with like on rapid charging, a garaged car holds 92.1% at 100,000 miles, a driveway car 91.2% and a street-parked car 89.8%, a 2.3 point range. Preconditioning before a winter rapid charge is worth a further 2.6 points, and only 42.8% of owners do it, which makes it the single cheapest improvement available to most drivers. Mean home charging session energy rises 23.2% between summer and winter, from 22.4 kWh to 27.6 kWh, which is the same cold-weather effect showing up as a bill rather than as degradation.

Longevity by segment and battery size#

Across the four multi-model segments, superminis reached 150,000 miles most often in 2026 at 8.9% and compact SUVs least often at 6.5%, despite superminis covering the lowest mean annual mileage at 6,840 and holding the third-lowest battery health at 100,000 miles.

This is a usage finding rather than a durability finding and the section says so in its first paragraph, because presenting it as durability would be wrong and would be spotted. The single-model estate segment leads the table outright at 13.4%, but one model is not a segment, and it is reported here rather than headlined for that reason. Among the segments with enough models to mean anything, the ordering is superminis at 8.9%, mid SUVs and saloons at 8.6%, large and premium at 7.5% and compact SUVs at 6.5%.

That ordering runs against the intuition, and the reason is worth setting out. Large and premium cars cover the highest mean annual mileage in the Panel at 10,180 miles, half as much again as superminis, and yet reach 150,000 miles less often. The explanation is age rather than annual distance: the supermini group contains the oldest models in the British parc, which have had the most years in which to accumulate distance, while the large and premium group contains sixteen models of which a large share are recent arrivals with no cohort old enough to have got anywhere near 150,000 miles. Segment figures here are unweighted means across the models in each segment, so a segment full of new models reads low whatever its cars are capable of.

The durability measure points a different way again, and that contrast is the useful part of this section. Holding mileage constant at 100,000 miles, the estate segment holds the most capacity at 89.4% and compact SUVs the most of the large segments at 88.5%, while large and premium cars hold the least at 86.4% and mid SUVs and saloons the second least at 86.8%, with superminis barely ahead of them at 86.9%. So the segment that reaches high mileage most often is close to the segment with the weakest batteries, and the segment with the best batteries reaches high mileage least often. Both statements are true, they are measuring different things, and quoting either one on its own would mislead.

Battery size behaves the same way. Capacity bands do not order cleanly on the share reaching 150,000 miles, because the band composition again reflects model age rather than pack engineering: 50 to 59 kWh cars lead at 10.1% and 80 to 89 kWh cars trail at 5.7%, with under-40 kWh cars at 8.7% in between. What does move consistently is the cycle count. The EV Cable Hub UK EV Lifecycle Panel 2026 found that covering 100,000 miles takes 892 full-equivalent cycles on a 40 kWh battery and 372 on a 100 kWh battery, a ratio of 2.4 to 1. A big pack does the same work in fewer cycles, and that is a real durability advantage that the share-reaching figures are too young to show yet.

A note for anyone planning to quote a segment figure. These rows are the unweighted mean of the models in each segment, so a segment containing sixteen models of which nine launched in the last three years will read low regardless of how durable those cars are. The segment table is a fair comparison of today's observed position and a poor guide to eventual outcomes, so the age-adjusted parc projection is published separately rather than being applied to segments where the cohorts are too thin to support it.

Battery replacement and repair#

3.2% of UK electric cars had a battery replaced by eight years old in 2026, and 2.4 percentage points of that was under warranty. The mean quoted cost of an out-of-warranty replacement was £8,240, and 38.6% of all battery work was completed at module level for a mean of £1,860.

This is the section people arrive at afraid, so it is answered with rates rather than with reassurance. Cumulative replacement runs at 0.6% by three years, 1.4% by five, 3.2% by eight, 5.8% by ten and 9.6% by twelve. The warranty split changes character over that period: at eight years, 2.4 of the 3.2 points are under warranty and 0.8 are not, while by twelve years only 3.1 of 9.6 points are covered. The mean state of health at replacement falls from 78.4% at three years to 66.4% at twelve, which is what you would expect if early replacements are warranty claims on faulty packs and late ones are ordinary wear.

The module-level finding is the genuinely under-reported one. Most people believe a failing electric car battery means a whole new pack. In the EV Cable Hub UK EV Lifecycle Panel 2026 it did not in 38.6% of cases: 24.6% of all battery work was a module-level repair under warranty and a further 14.0% was a module-level repair paid for by the owner, at a mean of £1,860 against £8,240 for a full pack. That is a mean saving of £6,380, and downtime of 11 to 14 days against 18 to 24. Smaller jobs are cheaper still: battery management system work quotes at a mean of £640 and coolant circuit work at £520.

Warranty terms are more uniform than most buyers assume, and the mileage cap is where claims fail. 62.2% of the 74 models carry 8 years or 100,000 miles, 13.5% carry 8 years or 125,000 miles and 8.1% carry 8 years or 160,000 miles, and every one of the six terms in the table uses a 70% health threshold. 6.4% of claims were rejected. The most common rejection reason was an exceeded mileage cap at 42.6% of rejections, and the second was no evidence of servicing at 24.8%. Both are avoidable, and the second one is avoidable for free.

Two further findings belong here. 46.2% of replacement packs fitted were remanufactured rather than new, at a mean state of health of 94.6%, which is better than the Panel mean at 100,000 miles, so a remanufactured pack is not a downgrade in capacity terms. And independent specialists carried out 58.4% of out-of-warranty work, at quotes averaging 41.2% below franchised dealer quotes. Mean time from fault report to completed work was 34 days, with 68.4% of owners offered a courtesy vehicle. Our EV battery replacement cost guide goes through the quotes in detail.

The replacement figures also need reading against what a replacement means in practice. A pack replaced under warranty at five years costs the owner nothing and takes a mean of 18 days. A pack replaced outside warranty at nine years costs a mean of £8,240 against a car that is often worth little more than that, which is why uneconomic battery repair accounts for 8.2% of parc exits rather than a larger share of repairs. The module-level route changes that arithmetic materially wherever it is available, and its availability is the single biggest open question in the used electric car market.

What actually goes wrong, other than the battery#

The 12V battery is the most common fault on a UK electric car, accounting for 31.4% of all faults reported in 2026. High voltage battery pack faults accounted for 4.8% and drive unit faults for 5.4%.

The fault distribution inverts the expected answer, which is why it is worth publishing plainly. The component that fails most often on an electric car is the same small lead-acid or lithium 12V battery that fails on a petrol car, at 8.6 faults per 100 vehicles per year and a mean repair cost of £142. It arrives early too, at a mean age of 3.2 years. The second most common fault is the charging port, flap or latch at 14.2% of faults and £286, and the third is the onboard charger unit at 9.6% and £1,240.

The high voltage battery pack, the failure the public associates with electric cars, is the eighth most common fault by share, at 4.8%. It is, however, comfortably the most expensive at a mean repair cost of £4,860, and the latest to arrive at a mean age of 6.8 years. The drive unit sits above it on frequency at 5.4% and below it on cost at £2,840. Between them the two big, expensive, distinctively electric components account for 10.2% of faults; the other 89.8% are the ordinary attrition of a car.

Fault rates rise with age in a way that is unremarkable and worth publishing anyway, because it is the table a used buyer needs. The EV Cable Hub UK EV Lifecycle Panel 2026 recorded 12.4 faults per 100 vehicles per year on cars under three years old, rising to 24.8 at three to five years, 38.6 at five to eight, 52.4 at eight to ten and 68.2 over ten. Mean annual maintenance spend follows the same curve, from £118 to £884. The share of owners reporting no fault at all in a year falls from 88.6% to 48.6% across the same span, and the share reporting three or more rises from 0.8% to 16.4%.

One caveat on this table specifically. Fault reporting is owner-reported, and it under-counts faults resolved under warranty without the owner recording them as a fault: a software fix applied during a service, for instance, or a part changed under a technical campaign. The relative ordering is reliable because the under-counting applies across systems; the absolute rates should be read as a floor.

Electric against petrol and diesel#

7.9% of UK electric cars have reached 150,000 miles in 2026, against 9.4% of petrol cars and 21.6% of diesels. Adjusted for the age of each fleet, the projected electric share is 18.4%, which places electric ahead of petrol and plug-in hybrid and behind full hybrid and diesel.

This is the comparison most easily misreported, so the age adjustment is given the same prominence as the raw figure rather than a footnote. The mean age of the British electric fleet is 3.8 years. The petrol fleet is 8.6 years and the diesel fleet 9.4. Comparing raw high-mileage survival across fleets that differ by six years tells you about the fleets, not about the cars. A three-year-old car has not failed to reach 150,000 miles; it has not had the chance.

Age-adjusted, the ranking is diesel at 24.8%, full hybrid at 19.6%, battery electric at 18.4%, plug-in hybrid at 13.4% and petrol at 11.2%. Diesel stays top for a reason that has nothing to do with engineering: diesel cars in Britain were bought disproportionately by high-mileage drivers, so a diesel fleet is a high-mileage fleet by selection. The EV Cable Hub UK EV Lifecycle Panel 2026 reports the electric figure as ahead of petrol by 7.2 percentage points and behind diesel by 6.4, and declines to make a stronger claim than the data supports.

Mean lifetime mileage tells the same story more conservatively, and here electric is genuinely behind: 78,400 miles against 94,200 for petrol, 132,600 for diesel and 108,400 for full hybrid. That gap is mostly the young-parc effect again, since a lifetime mileage figure cannot be observed for cars that have not yet left the road, but it is the number that will take longest to close and it should not be explained away.

Running costs are where the comparison is unambiguous. Mean annual maintenance spend is £286 for electric against £462 for petrol and £584 for diesel. Over 150,000 miles the total cost of energy, servicing, brakes and tyres, powertrain work, charging equipment, a home charge point and vehicle excise duty comes to £17,171 for an electric car against £34,685 for the petrol equivalent, or 11.4p a mile against 23.1p. Energy is where almost all of the difference sits, at £7,240 against £21,300. The electric column carries two costs the petrol column does not, a £1,100 home charge point and £186 of charging cables over the life of the car, and it still comes to half the total.

One caution on the cost comparison. It is built on the mean figures in this dataset and it assumes a car driven to 150,000 miles by an owner who charges mostly at home. A driver who cannot charge at home pays public rates for most of their energy, and the £7,240 energy line rises far enough to close a large part of the gap. The comparison is honest for the 78.6% of Panel owners with a dedicated home charge point and misleading for the 4.2% with no home charging at all, and both figures are published so a reader can place themselves.

Longevity and resale value#

Every additional percentage point of battery state of health at three years was worth 0.62 percentage points of retained value in 2026. Across the full best-against-worst charging comparison that is 14.2 percentage points of retained value, or £2,840 on a £20,000 car.

This is the section that turns the rest of the study into money, and the conversion rate is stable enough to state as a single coefficient. Every habit in the table converts battery health into retained value at between 0.61 and 0.64 points per point, so 0.62 is a fair working figure. That means the habits which protect a battery are not merely good practice, they are a return on behaviour the owner already controls and pays nothing for.

The largest single item is charging to 80% rather than 100% habitually, worth 5.2 percentage points of health against the 100% habit, 3.2 points of retained value and £640 on a £20,000 car, and 41.8% of owners already do it. Storing at 50 to 60% when the car is unused for more than a week is worth 5.0 points of health, 3.1 points of value and £620, and only 8.6% of owners do it, which makes it the largest unclaimed gain in the table. Rarely going below 20% state of charge is worth £400, preconditioning £320, a home cable that reaches with slack £340, a scheduled overnight window £280 and keeping rapid charging below 10% of energy £180.

The last row is the odd one and the most quotable. Recording state of health annually has no direct effect on the battery at all, and is worth 4.6 percentage points of retained value at sale, more than any habit that actually protects the pack. That is an information effect rather than a physical one: a documented degradation history removes the buyer's largest unknown, and buyers pay for that. Only 12.6% of owners keep such a record, and 58.4% have never checked their car's state of health at all.

The same relationship is visible from the other direction in the EV Depreciation Study 2026, where chemistry alone accounts for a 6.8 percentage point retention gap, and it shows up again in the Fastest Selling EVs Study 2026 in how quickly documented cars move. Read together with the depreciation figures, the practical conclusion is that battery care is worth roughly as much at resale as a full year of careful mileage management, and it costs nothing.

What this means if you own an electric car#

An owner who keeps rapid charging below 10% of their energy, charges to 80% for daily use, rarely goes below 20% and preconditions in winter reached 100,000 miles with 94.2% battery health in 2026. An owner doing none of those things reached the same mileage at 82.8%, a difference of 11.4 percentage points.

The gap compounds with distance rather than closing. At 50,000 miles best practice and worst practice are 5.7 points apart, at 100,000 miles 11.4 points and at 150,000 miles 14.8 points. Best practice reaches 150,000 miles in 14.6% of cases against 4.2% for worst practice, and carries a mean lifetime mileage of 94,800 against 62,400. That is a difference of 32,400 miles, roughly four extra years of average British driving out of the same car.

Not one of the eight habits costs money. All of them are settings, timing or where you park. And yet the EV Cable Hub UK EV Lifecycle Panel 2026 found only 22.4% of owners following four or more of them, and 18.6% following one or none. Owners doing four or more held 92.8% health at 100,000 miles against 86.1% for owners doing one or none, a 6.7 point difference in the real population rather than in the modelled extremes. Only 6.8% of Panel owners sit at or near full best practice, and 11.2% sit at or near worst practice.

The knowledge gap is wider than the behaviour gap and it is probably its cause. 58.4% of owners have never checked their car's battery state of health, 21.6% check annually or more often, 34.8% knew their battery warranty terms and 21.4% could state their car's chemistry. Most strikingly, 41.2% of owners believed an electric car battery needs replacing within ten years. The observed replacement rate by ten years in the Panel is 5.8%. That is a 35.4 percentage point gap between belief and measurement, and it is the single largest misconception this dataset can correct.

The ranking of the levers matters more than any individual number in it. Rapid charging share is worth up to 10.6 percentage points of battery capacity. Charge ceiling is worth 5.2. Charge floor is worth 3.2. Preconditioning is worth 2.6. Cable reach, through the plug-in frequency it enables, is worth 2.8. Home charging power is worth 1.4 once mileage is controlled for, which is to say worth nothing you should act on. The tools in the next section put your own numbers through the same coefficients.

There is one habit in the list that costs nothing and is not really about the battery at all. Recording state of health once a year, with a date and a reading, takes a few minutes and is worth more at resale than any single physical habit in this study. It also gives an owner the only reliable way to know whether their own car is tracking the Panel curve or falling below it, which is the difference between noticing a warranty-eligible fault at year six and noticing it at year nine.

Interactive tools#

A battery health projector built on the 74 model curves, a two-model longevity comparison, a searchable table of every figure on this page and a twenty-four item habit tracker that remembers where you got to. Everything runs in the browser.

Each tool draws on the tables above rather than on a separate dataset. With every behaviour input left at the 2026 Panel average, the projector returns exactly the figures published in Table 2 and Table 7 for the model selected, so the tool and the page cannot disagree.