EV Depreciation Study 2026: 74 Electric Cars Ranked on Value Retention and Pounds Lost

The 2026 headline findings#

A UK electric car retains 44.6% of its original list price after three years in 2026 and loses £21,840 in cash. The closest petrol equivalent retains 52.8% and loses £13,640, so the electric owner is £8,200 worse off on depreciation alone.

Every figure on this page is measured rather than forecast. The EV Cable Hub UK EV Lifecycle Panel 2026 calculates retention as the mean observed transaction price of a car at the target age, expressed against the original UK list price of the same model, trim, battery and drivetrain at the month it was first registered. That is the whole methodological difference between this study and most published depreciation figures, which are residual value projections written before the cars were sold. A projection describes what a market expects. A transaction describes what somebody actually paid.

Two headline numbers are published rather than one, and the gap between them matters. Weighted to the shape of the UK electric parc, three-year retention is 44.6%. Averaged flat across the 74 models with no weighting at all, it is 39.7%. The parc-weighted figure is higher because British buyers are concentrated in models that retain above the model average, and both are published so that anyone quoting one can see the other. The same split runs through the cash figures: £21,840 parc weighted against £28,360 unweighted, on mean list prices of £39,420 and £45,280 respectively.

Four structural drivers explain most of the ranking, and they are given here in order of measured strength. Original list price is first and by a distance, correlating with three-year retention at -0.71. Model generation age is next at -0.64: a car that is late in its generation loses value faster than an identical one early in a new one. Measured battery state of health is third at 0.58. Charging specification is fourth, with DC peak rate at 0.44 and onboard AC rating at 0.39. Nothing else the Panel tested reached 0.35, and acceleration reached 0.04, which is to say nothing at all.

The trend finding is more useful than the snapshot. The electric-to-petrol gap is closing, not widening. It stood at 19.4 percentage points in the 2024 edition of the Panel and stands at 8.2 points in 2026, a reduction of 11.2 points in two years. Used electric values fell 4.7% in the twelve months to June 2026 against 20.4% in the twelve months to June 2024. The direction of travel is unambiguous even though the level is still unflattering, and a study that reported only the level would be describing 2023.

The sections that follow take the dataset apart in order: the two centrepiece rankings first, then the shape of the curve at one, three and five years, then the cuts (segment, list price, battery size, chemistry, battery health, charging specification, mileage, colour, trim, region), then the petrol comparison, cost per mile, the 2026 price movement and what a buyer can actually control. Three sibling studies run off the same Panel and answer the questions this one does not: the Fastest Selling EVs Study 2026 on how long each model takes to sell, the Longest Lasting EVs Study 2026 on battery and drivetrain durability, and the Used EV Price Index 2026 on month-by-month price movement.

Three-year value retention, all 74 models ranked#

The Kia EV3 holds its value better than any other electric car in Britain, retaining 68.4% of its list price after three years in 2026. The Jaguar I-Pace is the worst at 21.6%, a spread of 46.8 percentage points.

Read the table one column at a time. Retention is the mean observed transaction price of a three-year-old example expressed as a percentage of the original UK list price of that same specification, drawn from the Transaction Panel. Percentage lost is simply its complement. Cash lost is the identical finding in pounds, printed in the same row on purpose so that nobody quoting this table has to do arithmetic to reach the number their readers care about. Value at three years is what the car was worth when it changed hands.

The pattern at the top is consistent and it is mostly about price and generation. Recent-generation models at moderate list prices dominate: the Kia EV3 at £32,995, the Skoda Elroq at £31,995 and the Renault 5 E-Tech at £22,995 take the first three places, and all three launched inside the last two years. The bottom of the table is the mirror image, filled with 2018 to 2020 premium cars that launched expensive and have since been overtaken by cheaper machines offering more range and faster charging.

The single strongest relationship in the whole study is between original list price and retention, at -0.71. Eight of the ten worst retainers launched at over £60,000, and the four worst (the Jaguar I-Pace, the Audi e-tron 55, the Porsche Taycan and the Mercedes EQC) launched at a mean of £80,723 between them. The EV Cable Hub UK EV Lifecycle Panel 2026 found nothing launched above £45,615 anywhere in the top quarter of this ranking.

Two cautions are worth stating before anyone lifts a row. Retention is measured against list price, not against what the first owner actually paid, so a model that was heavily discounted when new looks worse here than it was for the person who bought it at a discount. And eighteen of the 74 models have fewer than 36 months of transaction history, so their three-year figures rest on a shorter observation base than the rest. Both are set out in the methodology rather than buried in it.

The scatter below is the clearest single picture in the study. Each point is one model, positioned on its original list price and its three-year retention. The cluster of well-retaining cars sits almost entirely below £40,000, and above £70,000 there is no model retaining more than a third of its price. That is not a coincidence of this particular cohort; it is the arithmetic of a market where cheap electric cars have improved faster than expensive ones.

Pounds lost, all 74 models ranked#

The Dacia Spring loses less money than any other electric car in Britain, at £7,469 over three years in 2026. The Porsche Taycan loses £75,965, which is 10.2 times more, and the gap between the two is £68,496.

Percentages and pounds rank the same 74 cars very differently, and that difference is itself one of the more useful findings in this study. A model can retain a high percentage of a large number and still lose more cash than a model retaining a low percentage of a small one. Both measures are correct. They answer different questions, and a buyer choosing between segments needs the second one.

At the very top the two rankings agree. Eight of the ten best retainers by percentage are also among the ten smallest cash losses, because the cars that retain best are mostly cheap cars, and a cheap car cannot lose much money in absolute terms. Below the top ten the two lists pull apart badly, and the twenty models that move furthest are set out in their own table.

Four movements make the point on their own. The Nissan Leaf 40kWh sits 53rd on percentage retained and 13th on cash lost, a 40-place improvement, because 64.6% of £27,995 is £18,085 and that is a smaller cheque than most of the cars ranked above it write. The Mazda MX-30 moves 36 places the same way. In the other direction, the VW ID.Buzz is 20th on percentage and 54th on cash, 34 places worse, and the Kia EV9 is 29th on percentage and 60th on cash, 31 places worse. Both hold a respectable share of a large number.

The practical reading is straightforward. Quoting the percentage alone flatters expensive cars and penalises cheap ones, which is precisely backwards for anybody deciding how much money to put at risk. The EV Cable Hub UK EV Lifecycle Panel 2026 publishes both rankings in full for that reason, and the master data table at the end of this page lets either be searched directly.

One further pattern is visible in the cash ranking and is worth stating because it is not obvious from the percentages. The eleven largest cash losses in 2026 all belong to the large and premium segment, and every one of them launched above £62,000. Cash loss tracks list price more tightly than retention does, which is what makes it the more honest number to put in front of somebody comparing a supermini with a saloon.

One-year depreciation#

A UK electric car loses 33.2% of its value in its first year in 2026, an average of £13,080. That single year accounts for 59.9% of everything the car loses across its first three years.

Year one is where the money goes, and putting a precise share on it is what makes the finding usable. Of the £21,840 the average electric car loses in three years, £13,080 disappears in the first twelve months, £6,000 in the second and £2,760 in the third. On a monthly basis that is £1,090, then £500, then £230. The cost of ownership falls by more than three quarters between the first year and the third without anything about the car changing.

The mechanism has three parts and none of them is mysterious. The first is the drop from list price to the first retail transaction, which happens the moment the car is registered. The second is manufacturer discounting on new stock, which drags the nearly-new market down with it: a twelve-month-old car is competing directly against a brand new one carrying a factory contribution, and it can only win on price. The third is the loss of whatever premium attaches to being the current model year.

Some models resist it and by a wide margin. The Kia EV3 retains 80.4% after twelve months, the Renault 5 E-Tech 79.6% and the Skoda Elroq 78.8%. At the other end the same models that fail at three years fail at one, and the spread at twelve months is already 80.4% down to the low forties. The EV Cable Hub UK EV Lifecycle Panel 2026 found that the ordering established in year one is largely the ordering that survives to year five, which means the first-year figure is a reasonable early indicator rather than noise.

The curve table is the one to use for any age that is not a round number of years. It runs from three months to ten years and it is the basis for the calculators later on this page. The shape is worth noting: retention falls 11.6 points in the first three months, another 21.6 by twelve months, and then the gradient flattens sharply. Between 36 and 60 months the car loses 16.0 points; between 60 and 120 months it loses another 19.0, spread over twice as long.

The counterweight belongs here rather than buried later. Because retention is measured against list price and not against the transaction price when new, first-year loss is overstated for any model that was routinely discounted at the point of first sale. A car bought new at 12% off its list price has already absorbed part of that first-year fall before it left the forecourt. The affected models are noted in the methodology, and the distortion runs in one direction only: it never makes a car look better than it was.

Five-year depreciation#

A UK electric car retains 28.6% of its list price after five years in 2026, losing £28,160. The best five-year performer retains 46.2% and the worst retains 11.4%.

Five years is the horizon most private buyers actually own to, and it is badly under-published because it requires an observation window long enough to contain real five-year-old transactions. The EV Cable Hub UK EV Lifecycle Panel 2026 has that window for most of the field: five-year retention is observed for 50 of the 74 models and projected for the 24 that launched too recently to have a five-year population. Every row carries its basis in the last column so the two are never confused.

The distinction matters more than it might appear. A projected five-year figure is an extrapolation of a measured three-year curve, and extrapolations are wrong in predictable directions. They tend to be generous to new models, because a model that is fashionable at three years old is assumed to remain so. The projected rows in this table sit disproportionately at the top. Read the observed rows if you want the harder number.

The observed picture at five years is unflattering and should be stated plainly. The Jaguar I-Pace retains 11.4%, the Audi e-tron 55 12.1%, the Mercedes EQC 12.4% and the Porsche Taycan 12.6%. Those are 2018 to 2020 premium cars that have lost between 86% and 89% of their list price, and the cash figures behind them run from £62,016 to £86,002. No petrol equivalent in this dataset comes close to that.

The gradient between three and five years is gentler than most people expect, and it is remarkably consistent across the field. The average model loses a further 16.0 points of retention across those two years, and the spread around that average is narrow: the largest drop in the table is 22.2 points and the smallest is 10.2. Cars that have already fallen a long way fall more slowly from there, which is the ordinary arithmetic of depreciation against a fixed baseline rather than any kind of floor in the market.

For a buyer this is the most encouraging table on the page and it is worth saying so. A five-year-old electric car costs 71.4% less than it did new and has, on the Panel's own durability evidence, lost a mean of 9.8% of its battery capacity. Those two numbers are not proportionate to one another, and the gap between them is the entire case for buying used. The Longest Lasting EVs Study 2026 sets out the durability side in full.

The five-year table also settles an argument about early electric cars that comes up constantly. Cars from the 2018 to 2020 cohort have not held value, and the reason is not battery failure. Measured state of health at five years averages 90.2% across the parc. It is that they were expensive when new and have been overtaken on range, charging speed and efficiency by cars costing half as much. Depreciation on those models is a specification story, not a durability story, and the two get conflated more often than not.

Depreciation by segment#

Compact SUVs are the strongest electric cars for value retention in Britain, holding 43.7% after three years in 2026, just ahead of superminis at 42.4%. Large and premium electric cars are the weakest at 31.6%, and they lose 2.8 times the cash of the supermini segment.

Segment differences in this dataset are mostly price differences wearing a different label, and the honest way to present that is to show the segment figures and the price-band figures side by side and let the reader see how much of one is the other. The mean list price of the large and premium segment is £70,811 against £30,096 for superminis. Once that is on the page, a 12.1 point retention gap stops looking like a fact about body styles.

In cash the segment ranking is far starker than in percentage terms, and this is the version money desks should use. Superminis lose £17,667 over three years and large and premium cars lose £49,278. That is 2.8 times as much for a gap of 10.8 percentage points. At five years the same comparison is £22,239 against £58,248. The percentage spread understates the financial consequence by a wide margin, which is exactly the problem the pounds ranking earlier on this page exists to solve.

The within-segment table is the more actionable of the two, because nobody chooses a segment on depreciation grounds. Inside the compact SUV segment the Kia EV3 retains 68.4% and loses £10,426 while the Subaru Solterra retains 30.4% and loses £34,797. Inside mid SUV and saloon the Tesla Model 3 loses £19,035 and the Mercedes EQB loses £38,736. The EV Cable Hub UK EV Lifecycle Panel 2026 found the within-segment spread larger than the between-segment spread in every case, which means model choice matters more than segment choice.

The estate row carries a single model and is published as such rather than quietly dropped. One car is not a segment, the MG5 is both its best and its worst performer, and any comparison involving that row should be read as a comparison with one car. Publishing thin cells with their model count attached is more useful than suppressing them, provided the count is impossible to miss.

One counterintuitive result deserves a mention. Superminis do not lead this table despite being the cheapest segment, because the supermini field contains several early, short-range models (the Mazda MX-30 at 29.8% and the Honda e at 32.4% among them) that drag the segment mean down. Compact SUVs lead on the strength of a newer, more uniform field. Segment averages are hostage to the age composition of the cars inside them, and this is the clearest example of it in the study.

Depreciation by original list price#

Electric cars that launched below £25,000 retained 55.6% of their value after three years in 2026. Those launched at £75,000 and above retained 25.9%. Original list price is the strongest single predictor of retention in the whole study, correlating at -0.71.

This is the finding that explains most of the ranking, so it gets room and it gets tested openly. The gradient across the six price bands is close to monotonic: 55.6%, then 43.6%, 41.6%, 37.6%, 32.7% and 25.9%. Every step up in price costs retention, and the steps are large. In cash the same six bands run from £9,299 lost to £65,958 lost, a factor of seven.

The mechanism is not that expensive cars are worse. It is that the reference price they are measured against was set in a market that has since moved underneath them. An £80,000 electric car from 2021 is being valued today against 2026 machines that cost half as much and charge faster, and the used buyer prices the car it is rather than the car it was. Cheap electric cars have improved more, in absolute capability, than expensive ones over the same period, and the residual values reflect it.

The correlation table publishes every variable the Panel tested rather than only the ones that worked, which is the difference between research and marketing. Original list price leads at -0.71, days to sell in the used market follows at -0.68, and model generation age at -0.64. Battery state of health at three years comes in at 0.58 and DC peak charging rate at 0.44. At the bottom, kerb weight manages -0.14 and 0 to 62 mph time reaches 0.04, which is no relationship at all.

Two entries in that table are worth pausing on because they run against expectations. Usable battery capacity correlates negatively with retention at -0.18 (larger batteries are associated with slightly worse retention, not better), and WLTP range manages only 0.21. Both look wrong until price is accounted for, and both are taken apart properly in the battery size section that follows. The EV Cable Hub UK EV Lifecycle Panel 2026 publishes the raw correlation first and the controlled version second, in that order, so the working is visible.

The days-to-sell correlation at -0.68 is the strongest cross-study link in the programme. A model that sits unsold is a model being repriced downwards, and the two measures are close to interchangeable as early indicators. The Fastest Selling EVs Study 2026 ranks all 74 models on that measure directly.

Depreciation by battery size#

Electric cars with a 40 to 49 kWh battery retained 52.7% of their value after three years in 2026, the best of any capacity band. Cars with 90 kWh or more retained 31.1% and lost an average of £52,549.

The raw finding is real but it is mostly a price finding in disguise, and saying so is what separates this from the many published depreciation tables that control for nothing. Usable battery capacity correlates with original list price at 0.81. The 40 to 49 kWh band has a mean list price of £28,011 and the 90 kWh and above band a mean of £75,334. Any comparison between them is largely a comparison between a £28,000 car and a £75,000 one, and the price band section has already shown what that does.

Holding price band constant collapses the effect almost entirely. Across the capacity bands the uncontrolled spread is 14.4 percentage points; inside a single price band it is 3.6 points. Within the £25,000 to £34,999 band, retention runs from 44.2% to 45.6% across five different capacity bands, a range of 1.4 points from batteries differing by more than 30 kWh. Within the £45,000 to £54,999 band the range is 2.6 points.

What survives the control is small and it points the opposite way to the raw table. After price is held constant the mid-sized bands come out marginally ahead, at up to a point of advantage, and the 90 kWh and above band comes out 2.6 points behind. The EV Cable Hub UK EV Lifecycle Panel 2026 reads that as a weight and efficiency penalty rather than a capacity penalty: the very largest packs sit in the heaviest cars, and kerb weight is itself weakly negative at -0.14.

The WLTP range table tells a cleaner story than capacity does, because range is what a buyer actually experiences. Retention peaks in the 250 to 299 mile band at 46.6% and falls away in both directions: 31.2% below 150 miles and 38.4% above 400 miles. Under-range cars are penalised for being impractical and over-range cars are penalised for being expensive, and the sweet spot in the 2026 used market sits between them. That band also holds the largest share of the parc at 27.8%.

For anyone buying, the practical conclusion is that specifying the largest available battery is not a residual value strategy. It costs more new, it retains no better once price is accounted for, and above 400 miles of WLTP range it retains measurably worse while losing £42,480 in cash over three years. Range beyond what the driver uses is a comfort purchase, and the used market prices it as one.

Depreciation by battery chemistry#

Electric cars with LFP batteries retained 43.8% of their value after three years in 2026 against 39.1% for NMC. LFP cars also hold a 4.8 percentage point advantage in measured battery health at 100,000 miles, at 91.3% against 86.5%.

Chemistry is an under-published cut and a genuinely useful one, so it is worth defining the terms before using them. LFP is lithium iron phosphate: cheaper, heavier for a given capacity, and unusually tolerant of being charged to 100% every day. NMC is nickel manganese cobalt, the mainstream chemistry, which packs more energy into less weight and is more sensitive to being held at a high state of charge. NCA is nickel cobalt aluminium, a high-energy variant used in a small number of premium cars.

Chemistry and price are partly confounded here and the confounding is severe, because LFP is concentrated at the affordable end of the market. The ten LFP models in the Panel have a mean list price of £30,428 against £46,193 for the sixty NMC models. Uncontrolled, the LFP advantage over NMC is 4.7 percentage points. Controlled for price band it is 3.4 points. Most of the raw gap is real, but not all of it, and the controlled figure is the one to quote.

The two-chemistry cars are the most interesting row in the table and the smallest. Two models pair an LFP pack in the standard variant with NCA in the long-range one, and between them they retain 51.6% after three years, better than any single-chemistry group. That result rests on two models and should be read as such. It is published because suppressing a two-model row while publishing a ten-model one would be a choice about presentation rather than about evidence.

The durability evidence is what makes the retention finding coherent rather than a curiosity. At 100,000 miles LFP packs measure 91.3% of original capacity against 86.5% for NMC. That is a real difference in the thing buyers are most anxious about, and the used market appears to be pricing it, which is unusual given how little of the market knows the chemistry is different. The Longest Lasting EVs Study 2026 sets out the degradation curves by chemistry in full.

The awareness numbers are the counterweight and they are stark. In the Owner Odometer Survey 2026 only 21.4% of buyers knew which chemistry their own car used, and 9.6% considered it when buying. A market can price an attribute that almost nobody can name, because the price is set by the trade rather than by the private buyer, and the trade reads the auction data. LFP now accounts for 41.2% of models launched since 2023 against 6.8% of models launched before 2021, so this cut will matter more in each successive edition of the Panel.

The chemistry cut also has a practical implication for how a car is charged, which is where the two halves of the finding meet. LFP packs tolerate being charged to 100% routinely, and manufacturers generally recommend it. NMC and NCA packs prefer to sit lower for daily use. An owner who charges an NMC car the way an LFP car should be charged is measurably eroding the state of health that this study values at up to 34.8% of transaction price, which makes knowing the chemistry considerably more valuable than the 21.4% awareness figure suggests.

Battery health and what it is worth in pounds#

An electric car with a measured battery state of health above 95% sold for 11.0% more than an identical car with health between 85% and 89.9% in 2026. On a £20,000 car that is £2,200.

This is one of the three most linkable findings in the study, because it attaches a cash figure to something buyers argue about constantly and nobody quantifies. The premium is measured against the 90.0% to 94.9% band, which is the largest single band at 34.2% of transactions and is used as the baseline throughout. A car above 95% sells for 6.2% more than that baseline. A car between 85.0% and 89.9% sells for 4.8% less.

The gradient steepens sharply below 85% and that is where the real money is. The 80.0% to 84.9% band sells at 11.6% below baseline, the 75.0% to 79.9% band at 19.4% below, and cars below 75% at 28.6% below. On a £35,000 car that last step is £10,010. The full spread between the healthiest and least healthy band is 34.8% of transaction price, which is a larger effect than colour, trim, region and charging specification combined.

The disclosure finding is the one that will surprise people. Only 34.2% of listings state a battery health figure at all, which means the market is pricing an unknown for two thirds of cars, and it prices unknowns conservatively. Non-disclosure carries a 4.6% discount, worth £847 on the mean used electric car price of £18,420. A seller who has a healthy battery and does not say so is giving away most of the premium they earned by looking after it.

Disclosure is rising fast from a low base. It stood at 11.6% of listings in the 2024 edition of the Panel and 34.2% in 2026, roughly tripling in two years. The EV Cable Hub UK EV Lifecycle Panel 2026 also finds that a battery health certificate attached to a listing is worth 3.1 percentage points of retained value and cuts 17.2 days from the time the car takes to sell. The Fastest Selling EVs Study 2026 covers the days-to-sell side of the same disclosure effect in detail.

The reassuring half of this section is what the health figures themselves look like. Mean measured state of health is 93.1% at three years, 90.2% at five and 86.4% at eight. Annual decline runs at 2.3% through the first three years and 1.1% from year four to year eight, so the curve flattens rather than steepening. Only 3.6% of cars fall below 80% health before eight years. The market's anxiety about battery degradation is priced more heavily than the measured degradation justifies, and that gap is where a well-informed used buyer makes money.

One caveat on the source. These figures come from the Owner Odometer Survey 2026 and from listing disclosures, and listing disclosures are not independently verified in every case. Across matched records the mean gap between stated and measured health is 1.4 percentage points, with stated figures running high. That is small enough not to overturn the finding and large enough to be worth publishing.

Charging specification and residual value#

An 11 kW onboard AC charger was worth 7.2 percentage points of three-year retained value in 2026, £1,440 on a £20,000 car. DC charging at 150 kW or better was worth 8.4 points, or £1,680. Those are the two largest equipment effects the Panel measured.

Charging capability is now a priced attribute in the UK used market, and the premium is large enough to change a buying decision. The two variables need keeping apart, because they do different jobs. Onboard AC rating governs how fast the car charges at home or at a destination point. It is the number that matters on six nights out of seven. DC peak rate governs how fast it charges on a motorway, which matters on the days when it matters a great deal.

The raw band tables understate both effects and it is worth explaining why rather than quietly using the adjusted figures. Ranked by onboard AC rating alone, cars with 9.9 to 11 kW retain 40.6% and cars with 6.6 kW retain 37.8%, a 2.8 point gap. Ranked by DC peak alone, cars capped below 60 kW retain 37.4% while cars at 250 kW and above retain 36.1%, which reads as no advantage at all. Both rankings are contaminated by price: mean list price rises from £30,377 in the lowest DC band to £69,719 in the highest, and the price penalty swamps the capability premium.

The attribute table isolates the effect properly, comparing otherwise-similar cars, and it is the one to quote. An 11 kW onboard charger is worth 7.2 points and a 22 kW one 2.6. DC charging at 150 kW or better is worth 8.4 points. Vehicle-to-load capability is worth 3.8 points and, strikingly, cuts 7.8 days from the time the car takes to sell. A heat pump is worth 2.4 points. The EV Cable Hub UK EV Lifecycle Panel 2026 found every one of these effects positive and none of them trivial.

The mismatch between what buyers pay for and what owners use is the most interesting thing in this section. Used buyers pay a larger premium for rapid DC capability than for home AC capability (8.4 points against 7.2), yet owners in the Charging Behaviour Panel 2026 take 84.6% of their energy at home on AC. The attribute that gets used six days a week is priced below the attribute that gets used on holiday. That is a preference for reassurance over frequency, and it is entirely rational for anyone who has once needed a fast charge badly.

There is a practical caveat on the AC figure that sellers should understand. A car with an 11 kW onboard charger only delivers 11 kW if the supply and the cable can carry it, which for most British homes means a single-phase supply and a 7.4 kW ceiling regardless of what the car can accept. The premium is still real (it is paid on the car's capability, not on the buyer's installation), but the capability goes unused in a great many driveways. Our guide to EV charging cable amps explains how supply, cable and onboard charger interact.

The equipment finding at the end of the attribute table is the one with the shortest path to action. Including both charging cables at sale is worth 1.6 percentage points and cuts 18.4 days from the time the car takes to sell, the single largest days-to-sell effect of any attribute measured. Selling without any charging equipment costs 2.2 points and adds 9.8 days. If you are preparing a car for sale, the cables are worth more in the boot than on a shelf in the garage. Our range covers EV charging cables, granny chargers and V2L adapters.

One more attribute in that table deserves attention because it is growing fast. Vehicle-to-load capability appears on 36.5% of the parc and is worth 3.8 percentage points, but its effect on days to sell is 7.8 days. That is proportionally much larger than its price effect. That pattern, a modest price premium with a large liquidity premium, is the signature of a feature buyers actively search for rather than one they merely value once found.

Depreciation by mileage#

A three-year-old electric car with under 15,000 miles retained 54.2% of its value in 2026. The same car with over 75,000 miles retained 28.6%. Each additional 10,000 miles cost an average of 3.4 percentage points of retained value.

The mileage gradient is steep, continuous and easy to read off the table. Against the 25,000 to 34,999 mile baseline (which is where the median three-year-old car sits, and which retains exactly the parc average of 44.6%), a car under 15,000 miles carries a 9.6 point premium and a car over 75,000 miles a 16.0 point discount. In cash the same range runs from £18,060 lost to £28,160 lost.

Two points along that gradient steepen noticeably, and both are contractual rather than mechanical. The first is at 60,000 miles, where most manufacturer vehicle warranties have expired. The second is at 100,000 miles, where battery warranty terms typically lapse. The used market prices the loss of a warranty rather than any measured change in the car, and the steps are visible in the retention series either side of both thresholds.

The battery health column is what makes this table more than a mileage table. Measured state of health falls from 95.4% under 15,000 miles to 87.4% above 75,000, a loss of eight percentage points of capacity across the whole range. Retention falls 25.6 points across the same range. The market is applying roughly three points of price discount for every point of measured capacity loss, which is a considerably harsher exchange rate than the battery health section on its own would suggest.

That disproportion is the finding worth arguing with, and the argument is settled by durability evidence rather than by pricing evidence. A high-mileage electric car with 87.4% of its capacity intact has lost about 30 miles of a 240-mile range and has no cambelt, clutch, exhaust or timing chain to worry about. The Longest Lasting EVs Study 2026 follows high-mileage cars past both warranty thresholds and finds the failure rates that the pricing implies do not appear. High mileage is where the used electric market is cheapest relative to what the car will actually do.

The distribution matters for anyone interpreting the average. Only 8.4% of three-year-old electric cars have covered more than 60,000 miles, and 14.6% have covered under 15,000. The EV Cable Hub UK EV Lifecycle Panel 2026 records a mean annual mileage of 8,200 across the parc, which is materially below the equivalent petrol figure and reflects the fact that a large share of British electric cars are second household cars. That single fact does more to explain low-mileage electric stock than any assumption about how the cars are treated.

Depreciation by colour, trim and options#

Grey electric cars retained 46.2% of their value after three years in 2026, the best of any colour. Beige retained 36.4%, a 9.8 percentage point difference worth £3,863 on the average electric car.

Colour is a real effect and a moderately large one, and it is also partly a proxy for something else. The top four colours (grey, white, blue and black) account for 79.8% of the parc between them and separate by only 1.4 percentage points. Everything below them is a minority choice, and minority choices are correlated with particular trims, particular buyers and particular ordering patterns. Some of the beige penalty is a beige penalty and some of it is the company the colour keeps.

The cash column is the one to quote, because it is what the effect actually costs. Against the best colour, red costs £1,025, silver £1,340, yellow £2,286, purple £2,996 and beige £3,863 on the average electric car. Those are not trivial sums for a decision made in thirty seconds from a configurator, and they are entirely controllable at the point of ordering.

The options table carries the more useful finding and it is a short list of things that pay. A heat pump adds 2.4 percentage points, worth £946, against a typical option cost of £900. It is the only option in the table that roughly returns its purchase price. A tow bar adds 1.8 points, worth £710, against an £850 cost. Metallic paint adds 1.2 points for £650. Everything else loses money in resale terms, and some of it loses a great deal: a panoramic roof costs £1,200 new and adds £39, while larger alloy wheels cost £800 and subtract £315.

Trim behaves differently from options and rather more sharply. Performance and flagship trims retain 4.2 percentage points less than the model average (the largest negative in the table apart from service history), while mid trim retains 1.4 points more and high trim 0.6 points more. The EV Cable Hub UK EV Lifecycle Panel 2026 finds mid trim the best-retaining specification in almost every model line, which is the opposite of the usual advice to buy the top of the range and hold it.

Service history dwarfs everything else in this section and belongs in any summary of it. A full service history is worth 3.6 percentage points and no service history costs 9.4, a 13.0 point swing worth £5,125 on the average electric car. That is larger than colour, larger than every option combined, and it is the cheapest thing on the entire page to get right. It is also the item most owners of an electric car assume does not apply to them, on the grounds that there is less to service. The used market disagrees, decisively.

Depreciation by region#

Electric cars in South East England retained 45.8% of their value after three years in 2026, the strongest of the twelve UK regions. Northern Ireland was weakest at 41.6%, a gap of 4.2 percentage points worth £1,656 on the average electric car.

The regional spread is real but narrow, and the narrowness is itself the finding. Four points separate the best region from the worst, against 46.8 points between the best and worst model. Region is a rounding error next to model choice, and any regional story should say so before it says anything else. In cash the whole spread is £1,680, from £21,580 lost in the South East to £23,260 in Northern Ireland.

Four things drive what variation there is. Charge point density comes first: regions with dense public charging have deeper used electric demand. Off-street parking is second, because a household that can charge at home is a household that can own an electric car without arithmetic. Local stock mix is third, and a region whose used stock skews towards older premium cars will show weaker retention regardless of local demand. Household income is fourth and weakest.

The row-by-row picture gives every region a line of its own. South East England leads at 45.8%, with the East of England at 45.2% and the South West at 45.0%. The Midlands sit in the middle, at 44.8% and 44.6%. The North West holds 44.2%, Yorkshire and the Humber 43.8% and the North East 43.2%. Wales retains 42.8% and Scotland 42.4%. Greater London retains 42.1% and Northern Ireland 41.6%.

Greater London is the result that surprises people, and it is worth explaining rather than glossing. London accounts for 14.2% of UK electric car transactions, the second largest share of any region, and retains 42.1%, the second weakest. High transaction volume and weak retention sit together because London stock skews towards premium cars and towards fleet and lease returns entering the market in cohorts. Volume tells you where the cars are, not what they are worth.

The days-to-sell column shows a wider spread than retention does and the two move together. Cars take 39.8 days to sell in the South East and 52.4 days in Northern Ireland. That is a 31.7% difference against a 4.2 point difference in retention. The EV Cable Hub UK EV Lifecycle Panel 2026 attributes the mismatch to trade movement: cars are moved between regions to clear, which equalises price far more effectively than it equalises time on forecourt. Regional figures here are assigned by the transacting dealer's location, so a car sold across a border is counted in the selling region.

Electric against the equivalent petrol car#

A UK electric car retained 44.6% of its value after three years in 2026 against 52.8% for the closest petrol equivalent. In cash the electric car lost £21,840 against £13,640, a difference of £8,200, though the gap has narrowed from 19.4 percentage points in the 2024 edition of the Panel.

This is the most newsworthy comparison in the study and the one most likely to be misquoted, so the pairing logic comes first. Each electric model is matched to the closest petrol or hybrid equivalent by segment, body style and list price, over the same transaction window. Forty-two pairs qualified. Matching is on segment and price rather than on badge, which means some pairings are debatable, and the full pairing logic is published with the downloadable dataset rather than asserted.

The average favours petrol and there is no useful way to soften that. Across the 42 pairs the petrol or hybrid car retains more of its list price in 35 cases and the electric car in 7. The electric wins are concentrated in exactly the place the rest of this study would predict: recent, moderately priced models. The Kia EV3 beats the Sportage by 13.8 points, the Skoda Elroq beats the Karoq by 10.4 and the Renault 5 E-Tech beats the Clio by 8.8. The heaviest defeats are at the top of the market, where the Toyota bZ4X trails the RAV4 by 28.2 points and the Porsche Taycan trails the Panamera by 21.4.

The trend is the part that will still be true next year. The EV Cable Hub UK EV Lifecycle Panel 2026 records the gap at 8.2 percentage points against 19.4 points in the 2024 edition, a reduction of 11.2 points in two years. That is not a forecast, it is two measurements two years apart, and it is a considerably more informative statistic than the level. A study that reported only the level would be describing the used electric market as it was in 2023.

Depreciation on its own is not the whole answer, and a study that pretended otherwise would deserve to be picked apart. Over three years and 24,600 miles the electric car costs £1,404 in electricity against £4,182 in petrol, £846 in servicing against £1,584, and pays £324 more in insurance, £144 more in tyres, £300 for an amortised home charge point and £118 for charging equipment. Vehicle excise duty is now identical at £585.

Netted out, running costs recover £2,630 of the £8,200 depreciation gap. The three-year total is £28,063 for the electric car against £22,493 for the petrol equivalent, a difference of £5,570, or £1.141 per mile against £0.914. On these figures and at this mileage, total cost of ownership still favours petrol. The two things that reverse it are higher mileage, which multiplies the fuel advantage while leaving the depreciation gap unchanged, and buying used, which removes the depreciation gap almost entirely.

Two things about this comparison will change it materially in future editions, and both point the same way. The electric cohort being measured at three years old today launched in 2023, when electric list prices were at their peak relative to petrol equivalents. And the petrol cars they are matched against are being valued in a market where the supply of used petrol cars is contracting. Neither factor is permanent, and the 11.2 point narrowing already recorded is the first evidence of both unwinding.

Depreciation cost per mile#

Depreciation cost the average UK electric car owner 88.8p a mile over three years in 2026. The cheapest model cost 36.4p a mile and the most expensive 248.7p, which is 6.8 times more.

Cost per mile is the framing that makes depreciation comparable across segments, and it is the one fleet and leasing desks work in, so it gets its own ranked table. The calculation is deliberately transparent: three-year cash loss divided by the miles that car actually covers in three years, using each segment's observed mean annual mileage rather than a single flat assumption across the whole field.

Those mileages are published so the arithmetic can be checked. Superminis cover 6,840 miles a year, compact SUVs 8,120, estates 9,460, mid SUVs and saloons 9,240 and large and premium cars 10,180. Parc weighted the figure is 8,200 miles, or 24,600 over three years. The average electric car's £21,840 loss over 24,600 miles is 88.8p a mile, and every model row in the ranking uses its own segment's mileage in the same way.

Using segment mileage rather than a flat assumption changes the ranking materially and in the direction people expect least. Large and premium cars cover 49% more miles a year than superminis, which spreads their much larger cash loss over more miles and flatters them relative to a flat-mileage calculation. Even so they occupy the entire bottom of the table: the last eight places all belong to the large and premium segment, from 166.5p for the BMW iX to 248.7p for the Porsche Taycan.

The top of the table is where the useful comparison sits. The Dacia Spring costs 36.4p a mile in depreciation, the Kia EV3 42.8p, the Vauxhall Frontera Electric 43.1p and the Renault 5 E-Tech 43.5p. Those four are inside half the parc average. The EV Cable Hub UK EV Lifecycle Panel 2026 found no model launched above £45,615 anywhere in the cheapest third of this ranking, which is the price finding again in a different unit.

For context on the whole page, the equivalent petrol figure is 51.7p a mile against the electric 88.8p. That is the depreciation line alone. Set against it, the electricity used to cover those miles costs 5.7p a mile against 17.0p for petrol on the same three-year basis, so the running-cost advantage recovers roughly 11p a mile of a 37p disadvantage. Both halves of that sum belong in any honest cost-per-mile comparison, and one of them is routinely quoted without the other.

How depreciation moved through 2026#

Used electric car values fell 4.7% in the twelve months to June 2026, the smallest annual fall since 2022. Values rose in each of the three months to June 2026, the first three-month run of gains since May 2022.

The monthly series shows where the turn happened and it is not where most coverage placed it. Values were drifting gently through the second half of 2025, at losses of well under a percent a month, before a sharp two-month drop in November and December 2025 of 1.89% and 2.56%. January to March 2026 were shallow declines of 0.49%, 0.50% and 0.33%. April, May and June turned positive at 0.67%, 0.83% and 0.82%.

Three months of gains is not a recovery and this study will not call it one. Three data points cannot separate a turn from a seasonal pattern, and spring is reliably the strongest quarter in the used car year. What can be said with confidence is that the rate of decline has fallen sharply for three consecutive annual periods: -20.4% in the twelve months to June 2024, -12.6% to June 2025 and -4.7% to June 2026. Restraint about the monthly numbers is what will protect this section when the next edition publishes.

The cumulative picture is the necessary counterweight to the improving trend, and it is severe. Used electric values are down 46.2% from their May 2022 peak, which is £15,840 on a £30,000 basket. The mean used electric car was worth £34,260 in May 2022 and £18,420 in June 2026. Anyone who bought at the peak has taken a loss that four years of moderating declines has not begun to repair, and the index reading of 61.4 says exactly that.

The breadth of the recent move is more encouraging than its size. In the three months to June 2026, 71.6% of models saw their individual index rise and 28.4% saw it fall. A rally led by a handful of models would be a composition effect; one covering seven models in ten is a market-wide movement. The EV Cable Hub UK EV Lifecycle Panel 2026 tracks each model's index separately through the Listing Tracker for precisely this reason.

This section is the shortest bridge between this study and the Used EV Price Index 2026, which publishes the same index monthly rather than annually and carries the model-level series in full. Depreciation and price index are the same phenomenon measured on two clocks: this study measures a car against what it cost new, and the index measures the whole market against where it stood in a base month.

What this means if you are buying or selling#

Buying a three-year-old electric car instead of a new one avoided £21,840 of depreciation in 2026 and cost 55.4% less for the same vehicle. Choosing the best-retaining model in a segment rather than the worst saved between £17,810 and £51,241 over three years.

The two decisions above dominate everything else on this page and both are made before any other choice matters. Age at purchase is the larger of the two. A three-year-old car has already absorbed 55.4% of its list price in depreciation, and the buyer who takes it on absorbs 16.0 points over the following two years rather than the 55.4 the first owner took. Nothing else available to a buyer moves anything like that much money.

Model choice within a segment is the second lever and in some segments it is nearly as large. In large and premium the difference between the smallest and largest three-year cash loss is £51,241. In compact SUVs it is £25,290, in mid SUV and saloon £22,478 and in superminis £17,810. These are cars of comparable size and purpose bought by comparable people, and the only difference is which badge and which generation.

Below those two, the controllable decisions are smaller individually and meaningful in combination. Keeping a full service history is worth £5,125 against having none. Maintaining battery health above 95% is worth £4,494 against the 85 to 89.9% band. Avoiding beige and brown for grey, white or blue is worth £3,863. Choosing a model with 150 kW or better DC charging is worth £3,311 and one with an 11 kW onboard charger £2,838. Choosing mid trim over performance trim is worth £2,207.

The last three on the list are small enough to be dismissed individually and are not worth dismissing. Disclosing battery health at sale is worth £1,813, choosing LFP over NMC at the same price £1,340, and including both charging cables £631. The EV Cable Hub UK EV Lifecycle Panel 2026 finds a driver who takes all of the specification and condition decisions on the favourable side of this list protects several thousand pounds against one who takes none of them, on a car that is otherwise identical.

The share column in the decision table is the most quietly useful thing in this section, because it shows where the easy money is. Some of these decisions are already made by most buyers: 76.4% keep a full service history and 71.6% end up with an 11 kW onboard charger. Others are made by very few: 34.2% disclose battery health, 34.8% have 150 kW DC capability and 18.6% keep health above 95%. The decisions with the lowest take-up are, with almost no exceptions, the ones worth the most money, because an attribute that everybody has cannot command a premium for long.

Interactive tools#

Three calculators and a comparator built on the 2026 dataset, a searchable table of every figure on this page, and a twenty-five item depreciation-proofing 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, so every figure they return can be checked against the table it came from. The depreciation calculator and the new-against-used comparison share the same retention curve, so their answers always agree with one another and with Tables 2, 3 and 6.

The specification calculator adds point effects rather than compounding them, which is how the underlying tables are built. On the default £39,420 list price (the parc-weighted mean across the UK electric fleet), each individual attribute returns exactly the cash figure printed beside it in Table 20, Table 22 or Table 23.