Hot Weather EV Range Loss US 2026: 18,640 Summer Journeys, 61 Models and What Heat Actually Costs

The 2026 hot weather headline findings#

Electric vehicles lost 21.4% of their range above 95°F (35°C) and 26.8% above 100°F (38°C) in 2026. EV Cable Hub logged 18,640 US summer journeys between April and July 2026 and found air conditioning accounted for 61% of that loss, with battery thermal management taking a further 24%.

Heat loss is real, it is measurable, and it is not the mirror image of cold loss. The mechanism is different: cold weather range loss is dominated by cabin heating and by a battery that is chemically slower until it warms, and most of it is recovered the moment the weather turns. Hot weather loss is dominated by a compressor that has to reject heat from the cabin and from the pack at the same time, and it leaves a residue. The shape of the curve is different too, because cooling systems are sized for the common case and the extreme case sits outside what they were built for.

Air conditioning is the largest single cause at 61% of the loss, and it matters that it is the one load a driver will never switch off. A driver in February can put a coat on and turn the heater down; that trade exists and a meaningful number of people make it. A driver in a 105°F parking lot in Phoenix cannot make the equivalent trade, because the cabin is not merely uncomfortable without cooling, it is dangerous. Heat loss is therefore substantially less avoidable than cold loss, even though the headline percentage is smaller.

The most serious finding on this page is not about range at all. Vehicles based in the hottest states lost 4.1 percentage points more capacity over four years than identical vehicles in mild states, and that number does not come back in October. It is covered in full further down the page, alongside what the three mitigations are actually worth. Range loss costs a driver an afternoon of inconvenience and a few dollars. Capacity loss costs resale value.

Every figure here is measured against each vehicle's own baseline established from the same driver's journeys between 68°F and 77°F (20°C and 25°C), not against a published range rating. That is why these numbers differ from the rating-versus-real-world comparisons that circulate every summer, and it is why they are comparable across a fleet of 61 very different vehicles. The sections that follow take the picture apart: temperature bands first, then air conditioning, vehicles, pack thermal management, charging, cables, precooling, states, cost, degradation and behaviour.

Range loss by temperature band#

Range loss above 77°F (25°C) rises roughly linearly to 95°F (35°C) and then accelerates. EV Cable Hub's 2026 US data records 4.6% loss in the 77 to 80°F band, 16.8% between 90 and 95°F, and 32.1% above 110°F (43°C).

Three mechanisms stack as the thermometer climbs, and they do not arrive at the same time. The first is air conditioning demand, which rises with the gap between the cabin and the ambient air and again with solar load through the glass. That component is present from about 77°F and grows smoothly. The second is battery cooling, which begins drawing continuously once pack temperature passes the target window rather than cycling on and off, and which therefore appears as a step rather than a slope. The third arrives above roughly 100°F, when the pack cooling system can no longer hold the target during sustained high-power driving and the vehicle starts protecting itself by limiting what it will deliver.

That third mechanism is why the curve steepens instead of staying linear. A cooling system is specified for the conditions a manufacturer expects the car to spend most of its life in, with headroom for the occasional bad day. Above about 100°F the system is no longer working with headroom, it is working at its limit continuously, and the energy it draws rises faster than the temperature does. The energy-split table below shows the handover clearly: air conditioning accounts for 74% of the extra consumption in the 77 to 80°F band and only 51% above 110°F, not because the compressor is working less but because the pack has started demanding a much larger share.

The spread within each band is as informative as the mean. In the 95 to 100°F band the best journey recorded lost 11.4% and the worst lost 35.2%, a three-fold difference at the same ambient temperature. That is not measurement noise. It is the combined effect of vehicle, parking, journey length, precooling and cabin setpoint, and almost all of it is under some degree of driver control.

One caution on comparing these figures with anything else published. EV Cable Hub's 2026 US hot weather study measures journey-level consumption against each vehicle's own mild-weather baseline, taken from the same driver on the same roads. It is not a rating-versus-real-world comparison, which measures something different and produces larger numbers because it also captures the gap between the test cycle and ordinary driving. Both are legitimate. They are not interchangeable, and quoting one as the other is the most common error in summer range coverage.

The additional-energy table converts all of this into the unit drivers feel. Above 110°F a vehicle consumes 12.5 kWh more per 100 miles than in mild weather: $1.40 on an off-peak home rate, $6.00 at a DC fast charger.

What air conditioning actually draws#

Air conditioning drew a mean of 2.84 kW at 95°F (35°C) and peaked at 6.41 kW during cabin pull-down from a hot-soaked interior. EV Cable Hub's 2026 US testing found air conditioning accounts for 61% of all hot weather range loss, more than every other cause combined.

Almost every published air conditioning figure goes wrong in the same place: it fails to separate pull-down from steady state. A car cooling a 138°F cabin down to a liveable temperature is doing a completely different job from the same car holding an already cool cabin, and it draws more than double the power to do it. At 95°F the pull-down draw is 5.18 kW and the steady state draw is 2.84 kW. Quote either one as the figure for air conditioning and you are wrong for most of the journey.

Pull-down also takes longer than most drivers assume, and the duration rises faster than the temperature does. At 77°F a hot-soaked cabin reaches comfort in 3 minutes 10 seconds. At 110°F it takes 10 minutes 40 seconds, and at 115°F it takes 12 minutes 20 seconds, by which point the compressor has been running flat out for a fifth of an hour. This is the reason short journeys are so much worse than long ones in the heat, and it is quantified in the journey-length table below: a journey under two miles spends 88% of its extra energy on pull-down and loses 34.8% of its range, against 17.2% for a journey over a hundred miles.

The settings table is where a driver can actually act. Moving from fresh air to recirculation at the same 68°F setpoint cut the draw from 4.12 kW to 2.84 kW with no change at all in comfort rating, which makes it the single cheapest saving in this entire study. It costs nothing and no one notices. Raising the setpoint from 68°F to 75°F cut the draw to 1.86 kW and the comfort rating from 9.1 to 7.9, a real trade but a mild one. Ventilated seats at a 72°F setpoint held 74°F in the cabin at 1.98 kW and rated 8.8 for comfort, which is close to the best combination of cost and comfort in the whole table.

The windows-down question is the one readers arrive with, and EV Cable Hub's 2026 US testing settles it with a crossover rather than a verdict. At 35 mph the drag penalty of open windows is equivalent to 1.9 miles of range an hour against 8.8 miles for air conditioning, so windows win comfortably. At 55 mph it is 4.8 miles against 8.8, so windows still win on energy while the comfort rating has fallen to 4.8 out of 10. At 70 mph the penalty reaches 8.2 miles and the advantage has gone. The comfort-adjusted crossover sits at roughly 45 mph, which is why driving with the windows down on the highway is wrong in both directions.

Hot weather range loss by vehicle#

The gap between the best and worst electric vehicle in hot weather is 19.4 percentage points. In EV Cable Hub's 2026 US testing the strongest model lost 13.8% of its range above 95°F (35°C) and the weakest lost 33.2%.

Three engineering decisions separate the top of this table from the bottom, and none of them appear on a specification sheet a buyer is likely to read. The first and largest is whether the pack is liquid cooled or passively air cooled. The two passively cooled vehicles in the study lost 31.6% and 33.2% above 95°F, against a study mean of 21.4%, and they are the only two vehicles in the fleet whose loss above 100°F exceeds 40%. Passive cooling works acceptably in a mild climate and fails visibly in a hot one, and it is the single most consequential thing a buyer in Arizona or Nevada can check before signing.

The second is whether the air conditioning circuit is shared with the battery loop or separate. Vehicles with a dedicated battery cooling loop lost 17.8% above 95°F; vehicles sharing the cabin loop lost 20.6%. That is a 2.8 percentage point gap produced entirely by which compressor does what, and the mechanism is explained in the section below. It is the reason two vehicles with similar batteries, similar weight and similar mild-weather efficiency can sit four rows apart in this table.

The third is glass, and it is the finding that surprises people. A glass roof with no shade over a black interior held 152°F in the cabin after an hour parked in the sun at 95°F and needed 1.41 kWh to pull down. A solid roof over a light interior held 129°F and needed 0.94 kWh. That is a third of the pull-down energy saved by two specification choices made at the point of order, and it works out at 3.7 percentage points of range above 95°F between the two configurations. A retractable shade recovers roughly half the penalty of a glass roof, which makes the presence or absence of that shade a more useful thing to check than the roof material itself.

EV Cable Hub's 2026 US hot weather study makes no recommendation between these vehicles and the commentary here is deliberately neutral, because the table is the argument. Two readings are worth making explicit. Efficient vehicles are not automatically heat-resistant: the most efficient vehicle in mild conditions is not the most heat-resistant, and one of the worst percentage performers has a mild-weather efficiency well above the fleet mean. And percentage loss and absolute loss point in different directions for large vehicles, because a truck losing 21% of a small number of miles per kWh is in more practical trouble than a sedan losing 21% of a large one.

Parking recovers most of what specification took away. Shade rather than sun moved the same vehicle from 21.4% loss to 18.1%, and a garage moved it to 17.4%, larger effects than the difference between most pairs of vehicles in the table above.

Battery thermal management in extreme heat#

Passively cooled battery packs lost 31.6% of range above 95°F (35°C) against 17.8% for liquid-cooled packs with a dedicated loop, a gap of 13.8 percentage points. EV Cable Hub's 2026 testing found pack temperature exceeded the manufacturer's target window in 41.2% of journeys above 100°F on shared-loop vehicles.

Three cooling architectures appear in the 2026 US fleet and they behave very differently once ambient temperature exceeds what their designers assumed. A dedicated liquid loop gives the pack its own coolant circuit and its own claim on the compressor, so cabin demand and pack demand are met independently. A shared loop runs the cabin and the pack off the same refrigerant circuit, which is cheaper, lighter and adequate until both demands peak together. Refrigerant-direct cooling, present on only two vehicles in the study, puts refrigerant through the pack itself and produced the best figures of any architecture at 16.4% loss above 95°F. Passive air cooling has no active circuit at all and relies on ambient air the vehicle cannot make colder.

The shared-loop trade-off is the finding worth reporting. When the cabin needs cooling and the pack needs cooling from the same compressor at the same time, the vehicle has to choose, and almost every vehicle in the study resolves it in favour of the cabin. That is a defensible engineering decision (an occupant notices a hot cabin immediately and a hot pack not at all), but it means the pack spends longer above its target window on exactly the days it can least afford to. Shared-loop vehicles ran above target in 41.2% of journeys and hit a sustained power derate in 8.6% of them, against 18.4% and 2.1% for dedicated loops.

A thermal derate is not a warning light and most drivers never learn what it was. From the driver's seat it is a car that will not accelerate the way it did an hour ago, and that will not hold a set speed up a grade on the third hour of a hot highway drive. Mean derate depth in the 2026 US testing was 18%, and the mean time to onset at 110°F was 41 minutes of sustained highway running. It clears once the pack cools, which is why it is so rarely reported: by the time a driver reaches a service centre the car behaves normally.

The pack temperature table is the one to read if you want to understand why hot climates are hard on batteries rather than merely on range. Sixty minutes of highway driving at 110°F left the pack at a mean of 118°F and took 104 minutes to return to target after parking. A DC fast charge on top of that left it at 126°F with a 136-minute recovery. A passively cooled pack in the same conditions reached 138°F with a peak of 161°F and took 214 minutes to recover: nearly four hours, which in a desert summer means it frequently never recovers before the next drive. EV Cable Hub's 2026 heat charging test recorded that sequence repeatedly, and it is the direct mechanism behind the degradation figures later on the page.

Parking dominates recovery. Four hours in the sun at 110°F left the pack at 121°F; the same four hours in shade left it at 108°F, and in an 88°F garage at 91°F, within target.

DC fast charging in the heat#

DC fast charging power fell 31.6% above 100°F (38°C). EV Cable Hub's 2026 US testing recorded a mean peak of 118 kW from vehicles capable of 172 kW in mild conditions, and 20% to 80% charging times rose from 29 minutes to 44 minutes.

This is the finding that produces the most real-world frustration in hot states, because the three contributions stack rather than substitute. The driver arrives with a pack already heated by an hour of highway running in the heat. They plug into a cabinet whose own power electronics are derating because the air it rejects heat into is 105°F. And the session that results is slower than the one the trip planner assumed, which pushes the next leg later into the hottest part of the afternoon.

The vehicle-side contribution is the largest and the best documented. Pack temperature at arrival is what decides peak power, and the relationship is steep: a pack arriving in the 68 to 86°F optimal window pulled a mean of 172 kW and completed 20% to 80% in 29 minutes 10 seconds. A pack arriving between 104 and 113°F pulled 118 kW and took 42 minutes 20 seconds. Above 122°F it pulled 71 kW and took 68 minutes 40 seconds, a 135% penalty against optimal. That last band is small at 51 sessions and should be read as indicative, but the direction is unambiguous across all 1,196 DC fast sessions in the 2026 heat charging test.

The charger-side contribution is genuinely under-reported and it is the reason two drivers at the same site can have very different experiences. Below 86°F ambient, cabinets delivered 96.4% of rating and 2.1% were derating. Above 113°F they delivered 68.2% and 71.4% were derating, with 9.8% of sessions ended early by the charger rather than by the car. Cable cooling was active in 100% of sessions in that band, against 41.2% below 86°F. Not one of those figures is visible to the driver, who sees a slow session and blames the vehicle.

Precooling the pack before arrival is the mitigation and it is worth more than most drivers realise. A pack precooled before arriving at 100°F ambient pulled 152 kW and completed 20% to 80% in 32 minutes 50 seconds, a penalty of only 13% against optimal, against the 45% penalty a driver arriving at 104 to 113°F pack temperature would otherwise take. That is roughly twelve minutes recovered per stop for two kilowatt hours of energy spent on the way in, and on a road trip with three stops it compounds into a materially shorter day.

The vehicle-class table is what a driver planning a summer trip should read first, because architecture matters more than any of this in absolute terms. An 800V compact crossover went from 18 minutes 20 seconds in mild conditions to 31 minutes 10 seconds at 110°F. A passively cooled compact went from 51 minutes 40 seconds to 118 minutes 40 seconds. That is very nearly two hours for a 20% to 80% charge, a 130% penalty, and a figure that turns a road trip from slow into impractical.

Level 1 and Level 2 charging in the heat#

Level 2 home charging power fell 6.8% above 100°F (38°C) and Level 1 charging fell 9.4%. EV Cable Hub's 2026 US testing found that 22.1% of the delivered energy went into running the pack cooling system rather than into the battery during daytime charging above 104°F.

The important finding here is that home charging in extreme heat is not slower because the electronics are struggling. It is slower because the car is air conditioning its own battery while it charges, and that cooling energy comes out of the same supply. The delivered-against-rating column moves relatively little, from 97.1% below 77°F to 90.4% above 104°F. The energy-reaching-the-battery column moves much further, from 89.6% to 77.9%. The gap between those two columns is the cooling system, and it is the number that actually costs money.

That distinction has a consequence drivers can act on immediately, and it is independent of price. Charging at 4am on a 105°F day put 89.4% of delivered energy into the battery. Charging at 4pm on the same day put 77.6% in. Reaching 36 kWh at the battery therefore took 4.3 kWh of overhead at 4am and 10.4 kWh at 4pm, an extra 6.1 kWh purchased for exactly the same charge. At the 16.4 cent average rate that is $1.71 against $0.70 for the identical result, and the difference exists before any time-of-use tariff is applied.

Stack a time-of-use tariff on top and the two effects compound in the same direction, because summer air conditioning drives grid peaks in precisely the late-afternoon hours when a driver plugging in after work would otherwise charge. Overnight charging is cheaper per kWh and needs fewer kWh. EV Cable Hub's 2026 US owner survey found 71.4% of hot-state drivers had already shifted charging overnight against 31.8% in mild states, so the shift is happening fastest in the places where it bites hardest.

Shade and garages do most of the same work for drivers who cannot shift the time. Charging above 104°F in a garage delivered 96.1% of rating with 87.1% reaching the battery, close to the sub-77°F baseline. Charging above 104°F in the open delivered 90.4% with 77.9% reaching the battery. Session time for 36 kWh at Level 2 ran 5 hours 26 minutes in the garage against 6 hours 4 minutes in the open, a 38-minute difference produced by nothing but where the car was parked.

Level 1 charging degrades slightly faster in relative terms, from 96.8% below 77°F to 88.1% above 104°F, and it starts from a much lower absolute rate, so a Level 1 charger in a hot climate can lose ground overnight against a day's driving. Readers sizing a circuit or a portable unit will find the practical detail in our guides to charging amps and 16A against 32A cables. US Level 1 and Level 2 map onto the Mode 2 and Mode 3 terminology used in Europe, which our charging modes guide and portable charger guide explain, and the equipment itself sits in the portable charger range.

Charging cables in extreme heat#

Charging cable jacket temperature reached 147°F (64°C) lying on asphalt at 100°F (38°C) ambient, and 168°F (76°C) in direct sun on dark asphalt. EV Cable Hub's 2026 bench programme found conductor resistance rises 18.4% between 68°F and 158°F, which is a real and measurable loss of delivered power.

A cable lying on a surface in the sun is not at ambient temperature and almost nobody measures what it actually reaches. On dark asphalt in direct sun at 100°F ambient the jacket ran at a mean of 168°F with a peak of 179°F, and the conductor inside it reached 184°F while carrying 32A. That is 68°F above the air temperature a driver would quote. Coiled on the same surface it ran hotter still, at a mean of 174°F and a peak of 188°F, because a coil traps its own heat as well as absorbing the sun's. That is the practical reason to uncoil a cable fully before charging, and a point our coiled against straight cable comparison covers in more detail.

Hot conductors carry more resistance, and more resistance means less power reaching the car. Across the 68 cables in the 2026 heat bench programme, resistance rose to 118.4% of its 68°F value by 158°F and to 126.4% by 194°F. On a 7.7 kW circuit that took delivered power from 7.41 kW to 7.29 kW, with the difference dissipated as heat in the cable itself: 44 W at 68°F rising to 56 W at 194°F. The magnitude is small and should be reported as small: it is roughly a 1.6% power loss across a 126°F temperature swing, not a reason to change any purchasing decision. It is, however, real, measurable and never disclosed at the point of use.

Jacket softening is the inverse of the winter stiffness problem and it produces its own failure modes. PVC began softening at 141°F, deformed 18.6% under its own weight at 158°F, adhered to asphalt at 44.8% and lost 52% of its abrasion resistance. TPU began softening at 171°F, deformed 2.1%, adhered at 4.4% and lost 14%. Rubber compound and silicone hybrid jackets did better still. On a 168°F asphalt surface a PVC-jacketed cable is a soft object stuck to the ground, and the abrasion figure is what matters when it is then driven over. In the 2026 US owner survey, 6.8% of drivers reported exactly that damage. Length matters here too, because a longer run means more of the cable lying on the hot surface, which our long cable guide takes up.

Connectors are the other thermal limit and they are the one that stops a charge. At 77°F ambient a connector body reached 104°F after two hours at 32A and 0.2% of sessions triggered a thermal derate. At 113°F ambient it reached 154°F and 14.8% of sessions derated, at a mean depth of 24%. Leaving that same connector in direct sun took it to 167°F and 28.4% of sessions; shading it took it to 146°F and 9.1%. Shading the connector alone cut derate incidence by two thirds, which is the cheapest intervention on this page.

The driver-reported problems table is the human end of the same data. 38.4% of US drivers reported a cable too hot to handle comfortably, 26.8% a connector too hot to grip when unplugging, and 24.6% jacket fading or chalking from UV exposure. Only 31.2% reported no summer cable problems at all. Storage is most of the answer: a cable in a closed trunk parked in the sun reached 156°F, against 118°F in a shaded trunk and 92°F in a garage. Our charging cable range lists jacket material on every product for readers who want to check it against the table above.

Precooling, and what it saves#

Precooling while plugged in cut hot weather range loss by 6.2 percentage points in 2026, from 21.4% to 15.2% on the same vehicles and journeys. Only 46.8% of US drivers precooled regularly.

There are two kinds of precooling and most drivers only know about one. Cabin precooling cools the interior before departure so the compressor is not doing pull-down work on the move. Pack precooling brings the battery down towards its optimal window before a DC fast charging stop, and it is the more valuable of the two by a wide margin. Yet EV Cable Hub's 2026 US owner survey found only 22.1% of drivers knew their vehicle could do it and only 11.4% had ever used it before a charging stop.

The plugged-in distinction is the whole argument. Precooling from the wall for ten minutes cost 0.86 kWh of grid energy, took nothing from the pack and cut loss at 100°F from 26.8% to 20.6%. Precooling for the same ten minutes unplugged took the identical 0.86 kWh out of the pack and cut loss only to 25.4%, because the energy spent came out of the range it was meant to protect. Plugged in, precooling is close to free. Unplugged, it is worth roughly a fifth as much. The common claim that precooling wastes energy is true only of the unplugged case and is false in the case almost everyone is actually in.

Duration has a clear point of diminishing returns. Five minutes plugged in took loss from 26.8% to 22.4%. Ten minutes took it to 20.6%. Fifteen minutes took it only to 20.1% for half again as much wall energy, and the cabin had already reached comfort by ten. Ten minutes is the number to publish. Stacking a windshield sunshade on top of ten minutes of plugged-in precooling produced the best result in the table at 18.4%, against 24.1% for the sunshade alone.

Precooling the pack before a DC fast charging stop is the finding almost nobody acts on. It costs 2.11 kWh and recovers roughly twelve minutes at the charger, because a pack arriving in or near its optimal window pulls 152 kW rather than the 118 kW a hot arrival manages. On a hot-weather road trip with three stops that is well over half an hour of the day recovered for about six kilowatt hours of energy, and it also spares the pack the high-temperature fast charging that the degradation section shows to be the most damaging combination in the study.

Why so much of this value is left on the table is a question of interface: 52.4% of drivers who precool use the app rather than a schedule, and an app is a decision that has to be made every hot day.

• Precool on most hot days: 46.8%
• Precool occasionally: 28.4%
• Never precool: 24.8%
• Know their vehicle can precool the pack separately: 22.1%
• Precool before a DC fast charging stop: 11.4%
• Precool while unplugged: 34.6%
• Use a scheduled departure time: 31.8%
• Use the app rather than a schedule: 52.4%
• Use a windshield sunshade regularly: 61.2%
• Park in shade whenever available: 78.4%
• Have a garage available: 42.6%
• Use the garage for the EV: 71.1% of those who have one

State by state heat exposure#

Arizona drivers lost 19.8% of their range across the 2026 summer against 5.0% in Washington State, a gap of 14.8 percentage points. EV Cable Hub's 2026 data covers 44 states and attributes almost all of the variation to days above 90°F and to overnight low temperatures rather than to daytime peaks alone.

The table below is sortable on every column, so a state outlet can find its own row in seconds and a reader can rank the whole list by days above 90°F, by overnight low or by loss on the hottest day rather than by the mean. Two of the rows are split by geography rather than by state line, with California and Oregon each appearing as an inland and a coastal entry, because treating either as a single climate would misrepresent both halves badly. Coastal Oregon lost 3.4% and inland Oregon 8.1%; coastal California lost 4.1% and inland California 15.6%.

The counterintuitive finding is that overnight lows matter as much as daytime highs, and the correlation table proves it: days above 90°F correlates at 0.94 with mean summer range loss, mean overnight low at 0.88, and mean summer high at only 0.86. A pack that never gets below 82°F overnight starts every morning already warm, spends the day climbing from a higher base and never gets a full recovery window. A pack in a high desert that drops to 60°F overnight starts each day inside its target window even when the afternoon reaches 95°F.

That mechanism is why the desert Southwest performs worse than the humid Southeast at comparable daytime peaks. New Mexico records a 95°F mean summer high, higher than Georgia's 92°F, but a 66°F overnight low against Georgia's 72°F, and New Mexico loses 13.4% against Georgia's 14.1%. Arizona is the extreme case on both measures at once: 148 days above 90°F, a 106°F mean high and an 82°F mean overnight low, and it tops the table by a clear margin at 19.8% with 33.4% on its hottest day.

Two predictors are worth naming for what they are not. Mean relative humidity correlates at only 0.21 with state-level loss, which does not make humidity unimportant (it is worth 3.4 percentage points at a fixed temperature), but it does not vary enough between states to drive the ranking. The share of drivers with garage parking correlates at -0.61, the strongest driver-side predictor on the list.

Six states are not represented and are excluded rather than estimated. State figures reflect the summer of 2026 specifically, which ran hotter than the ten-year mean across the Southwest.

Humidity, sun load and the things nobody measures#

High humidity added 3.4 percentage points to range loss at the same ambient temperature in 2026, because the air conditioning has to remove moisture as well as heat. Direct sun load added a further 2.8 percentage points on identical journeys.

These are the second-order variables that make a Houston summer feel different from a Phoenix summer at the same thermometer reading, and quantifying them is genuinely new. An air conditioning system does two jobs at once: it lowers the temperature of the air, and it condenses moisture out of it. The first is sensible cooling and the second is latent cooling, and only the first is what most people mean by air conditioning. At 25% relative humidity, 6.1% of the compressor's energy goes on dehumidification. At 75% relative humidity it is 31.4%, five times as much work for the same temperature drop.

Sun load through glass is the second variable and it is larger than most drivers would guess. A standard glass roof admitted 412 W of solar energy into the cabin, against 148 W with a shade closed and 61 W through a solid roof. The windshield admitted 386 W without a shade and 94 W with a reflective one. Those are continuous loads the compressor has to remove on top of everything else, and they are the mechanism behind the parked-cabin temperatures in the vehicle section: 412 W arriving through a roof for four hours is a great deal of heat to put back out.

The same effect shows up while driving. An identical journey run into a low sun with full air conditioning lost 22.6% of range; the same route after dark at the same ambient temperature lost 17.4%. That 5.2 percentage point difference is sun load and nothing else, and it is why late-afternoon westbound driving is the worst combination in a hot climate: the ambient temperature is near its peak, the sun is low enough to come in through the windscreen rather than the roof, and the pack has had all day to warm up.

Surfaces are the third variable and here the finding is that it barely matters, which is worth publishing as clearly as the ones that do. Asphalt in direct sun reached 141°F at 95°F ambient and 162°F at 110°F, against 138°F for concrete at 110°F. Hot air raised tyre pressure by 4.6 psi from a cool morning to a hot afternoon, which is worth 0.8% of efficiency on its own, while hot pavement raised rolling resistance by 2.1%. The net effect of hot surfaces on range was a loss of 1.3%. Two real effects, pointing in opposite directions, and a small residual. EV Cable Hub's 2026 US hot weather study reports it as small because it is small.

• Range loss at 95°F and 25% relative humidity: 19.8%
• Range loss at 95°F and 75% relative humidity: 23.2%
• Humidity penalty: 3.4 percentage points
• Share of air conditioning energy spent on dehumidification at 75% humidity: 31.4%
• The same share at 25% humidity: 6.1%
• Range loss on an identical journey in full sun against overcast at 95°F: 2.8 points higher
• Sun load through a standard glass roof: 412 W
• Sun load through a glass roof with a shade closed: 148 W
• Sun load through a solid roof: 61 W
• Windshield sun load, no shade: 386 W
• Windshield sun load with a reflective shade: 94 W
• Asphalt surface temperature at 95°F ambient, direct sun: 141°F
• Asphalt surface temperature at 110°F ambient, direct sun: 162°F
• Concrete surface temperature at 110°F ambient: 138°F
• Tyre pressure rise from cold morning to hot afternoon: 4.6 psi
• Range effect of that pressure rise: 0.8% better efficiency
• Rolling resistance change on hot asphalt against cool: 2.1% higher
• Net effect of hot surfaces on range: 1.3% worse
• Range loss driving into a low sun with full air conditioning: 22.6%
• Range loss on the same route after dark at the same ambient: 17.4%
• Difference attributable to sun load: 5.2 percentage points

Highway driving in extreme heat#

Highway driving at 75 mph in 105°F (41°C) heat produced a mean efficiency of 2.41 mi/kWh against 3.18 mi/kWh in mild conditions, a loss of 24.2%. EV Cable Hub's 2026 US testing found 8.6% of vehicles entered a sustained power derate during highway driving above 100°F.

Winter and summer invert here, and the inversion is the point of the section. In cold weather a highway journey is the most efficient kind of driving an electric vehicle does, because the cabin heater is a fixed load spread over more miles and the pack warms itself usefully as it works. In extreme heat a highway journey is where the thermal system is under most stress, because motor and inverter heat, pack heat and cabin cooling demand all peak at the same time and the only thing available to reject that heat into is 105°F air.

Speed is the lever, and it is a bigger one in the heat than in mild conditions. At 105°F, dropping from 75 mph to 65 mph moved efficiency from 2.41 to 2.74 mi/kWh and range loss from 24.2% to 20.1%. Dropping to 55 mph moved it to 3.04 mi/kWh and 16.4%. Ten miles an hour is worth 4.1 percentage points, which is more than the difference between a dedicated cooling loop and a shared one. Aerodynamic drag accounts for most of it, but the thermal contribution is real: a slower car generates less motor heat and asks less of a cooling system already at its limit.

Derating is the tail risk and it is concentrated at the top of the temperature range. Above 100°F, 8.6% of highway journeys entered a sustained power derate; above 110°F it was 21.4%, at a mean depth of 18% and a mean time to onset of 41 minutes. That last figure is the useful one for anyone planning a summer drive: the first forty minutes are normal, and the problem arrives on the second hour. It is also why the stop-and-go figures later in this page look so different. Low speed means low motor load, and derating is essentially a highway phenomenon.

Towing is severe and almost entirely undocumented anywhere. Towing 3,500 lb cost 47.8% of range at 95°F against 38.4% in mild conditions, an additional heat penalty of 9.4 percentage points on top of an already large towing penalty. A roof cargo box cost 18.1% at 75 mph in heat against 6.4% for crossbars alone. Both are cases where added drag and added thermal load compound, and both are worth planning around rather than discovering at the second charging stop.

The compounding shows up in trip structure. EV Cable Hub's 2026 US hot weather study recorded a mean highway leg of 118 miles in hot states against 152 in mild ones.

• Efficiency at 75 mph, mild conditions: 3.18 mi/kWh
• Efficiency at 75 mph, 95°F: 2.68 mi/kWh
• Efficiency at 75 mph, 105°F: 2.41 mi/kWh
• Efficiency at 75 mph, 110°F: 2.28 mi/kWh
• Efficiency at 65 mph, 105°F: 2.74 mi/kWh
• Efficiency at 55 mph, 105°F: 3.04 mi/kWh
• Range loss at 75 mph in 105°F: 24.2%
• Range loss at 65 mph in 105°F: 20.1%
• Range loss at 55 mph in 105°F: 16.4%
• Vehicles entering sustained power derate above 100°F on highway: 8.6%
• Vehicles entering derate above 110°F on highway: 21.4%
• Mean derate depth when triggered: 18%
• Mean time to derate onset at 110°F: 41 minutes
• Range loss towing 3,500 lb at 95°F: 47.8%
• Range loss towing 3,500 lb in mild conditions: 38.4%
• Additional heat penalty when towing: 9.4 percentage points
• Range loss with a roof cargo box at 75 mph in heat: 18.1%
• Range loss with roof crossbars only: 6.4%
• Mean highway leg length between charges, hot states: 118 miles
• Mean highway leg length between charges, mild states: 152 miles
• Extra charging stops on a 500-mile trip in extreme heat: 1.4

What summer heat costs in money#

Summer heat added $214 to the average US EV driver's annual energy bill in 2026. That rises to $486 for drivers in Arizona who rely on DC fast charging, and falls to $61 for drivers in coastal states charging overnight at off-peak rates.

The arithmetic is shown openly so it can be checked. Two separate mechanisms drive the cost, and they are reported separately in the profile table because they respond to completely different actions. The first is volume: a hot summer simply consumes more kilowatt hours for the same miles, 468 kWh more for a typical 12,000-mile hot-state driver. The second is price: summer air conditioning drives grid peaks in exactly the late-afternoon and early-evening hours when a driver arriving home would otherwise plug in, so more of those kilowatt hours get bought at peak rates.

Separating them matters because the second mechanism is often larger than the first and is entirely avoidable. A 12,000-mile hot-state driver on a time-of-use tariff paid $52 in extra volume and $162 in extra peak pricing, a total of $214. The same driver, same miles, same 468 kWh, on a flat rate paid $77 and nothing in peak pricing, a total of $77. The tariff is not the villain, since a time-of-use tariff is cheaper overall for a driver who charges overnight, but a driver on a time-of-use tariff who charges at 6pm in August is taking the worst of both structures.

At the top of the table the numbers get serious. An 18,000-mile Arizona driver charging predominantly at DC fast chargers paid $342 in extra volume and $144 in extra peak pricing, a total of $486 for the summer, and a 20,000-mile Texas fleet vehicle paid $334. Those drivers are disproportionately the ones without home charging.

The cost-per-mile table puts the summer uplift on a common basis and produces a consistent result: every electric charging source rises by close to 27% between mild and summer conditions, because the uplift is a consumption effect rather than a price effect and it applies proportionally whatever the rate. Home off-peak went from 3.04 to 3.86 cents a mile. DC fast at 48 cents went from 13.04 to 16.55 cents. Workplace charging offered free stayed at zero in both columns, which is the only row in the table immune to summer.

The gasoline comparison is the honest counterweight and it is not flattering everywhere. Gasoline rises only 5.1% in summer, because a combustion engine's air conditioning is a much smaller fraction of a much larger energy flow. A 25.4 mpg gasoline car cost 13.41 cents a mile in summer 2026. An electric vehicle charged at a 48-cent DC fast charger cost 16.55 cents. For a driver with no home charging in a hot state, the running-cost case against an average gasoline car does not hold through the summer. And against an efficient 32.1 mpg car at 10.61 cents a mile it does not hold against public Level 2 either. EV Cable Hub's 2026 US hot weather study publishes that because it is what the data shows.

Heat and long-term battery degradation#

Vehicles based in the hottest states lost 4.1 percentage points more capacity over four years than identical vehicles in mild states. EV Cable Hub's 2026 US analysis put mean four-year capacity retention at 88.4% in the hottest states against 92.5% in mild states.

This is the most serious hot weather finding in the study and the one with the longest life, because it does not affect one afternoon's driving, it affects resale value, warranty position and whether a used vehicle from a hot state is worth the same as one from a mild state. It should be read carefully. The analysis compares vehicles of matched age and mileage across climates rather than running a controlled long-term trial, so it establishes a strong association rather than proving causation, and that qualification is stated here as well as in the methodology.

Three factors stack, and the largest by far is architecture rather than climate. Passively cooled packs in hot states retained 79.6% of capacity over four years against 89.8% for liquid-cooled packs in the same states, a 10.2 percentage point gap, two and a half times the size of the hot-state-against-mild-state gap itself. A liquid-cooled pack in Arizona did better than a passively cooled pack anywhere warm. If there is one purchasing conclusion in this entire page, it is that one.

The other two factors are behavioural and both are actionable. Vehicles held above 80% state of charge in heat retained 86.4% against 90.8% for vehicles held between 50% and 70%, a 4.4 point difference produced by nothing more than where the driver stops charging. High state of charge and high temperature together are the combination that ages a cell fastest, and a full pack sitting in a hot parking lot all afternoon is precisely that combination. And vehicles DC fast charging more than half the time in hot states retained 85.9% against 90.2% for those doing so less than a tenth of the time, a 4.3 point difference. A fast charge into an already hot pack is the most thermally aggressive thing an electric vehicle routinely does.

Parking is the third lever and the easiest. Vehicles garaged overnight in hot states retained 90.6% against 87.1% for vehicles parked outside, a 3.5 point gap. Taken together, garage parking, moderate state of charge and limited DC fast charging were worth a combined 7.9 percentage points of four-year retention in EV Cable Hub's 2026 US analysis. That is nearly twice the size of the raw climate penalty, which means a careful driver in a hot state can retain more capacity than a careless driver in a mild one.

The consequences are worth stating in proportion rather than in alarm. Annual degradation ran at 2.9% in the hottest states, 1.9% in mild states and 1.6% in the coolest. Only 1.4% of vehicles in hot states reached the 70% capacity threshold within their warranty terms, so this is a value question rather than a failure question. The mean resale effect of 4.1 points of extra capacity loss was $1,180, which is roughly five and a half years of the $214 annual energy penalty arriving as a single deduction at trade-in.

• Four-year capacity retention, hottest states: 88.4%
• Four-year capacity retention, mild states: 92.5%
• Gap: 4.1 percentage points
• Four-year retention, passively cooled packs in hot states: 79.6%
• Four-year retention, liquid-cooled packs in hot states: 89.8%
• Cooling architecture effect: 10.2 percentage points
• Retention for vehicles garaged overnight in hot states: 90.6%
• Retention for vehicles parked outside in hot states: 87.1%
• Garage effect: 3.5 percentage points
• Retention for vehicles held above 80% state of charge in heat: 86.4%
• Retention for vehicles held at 50 to 70% in heat: 90.8%
• State of charge effect: 4.4 percentage points
• Retention for vehicles DC fast charging over 50% of the time in hot states: 85.9%
• Retention for vehicles DC fast charging under 10% of the time in hot states: 90.2%
• DC fast charging effect: 4.3 percentage points
• Annual degradation rate, hottest states: 2.9%
• Annual degradation rate, mild states: 1.9%
• Annual degradation rate, coolest states: 1.6%
• Vehicles in hot states reaching 70% capacity within warranty terms: 1.4%
• Mean resale value effect of 4.1 points of extra capacity loss: $1,180
• Combined mitigation value of garage, moderate state of charge and limited DC fast charging: 7.9 percentage points of retention

Road trips in extreme heat#

A 500-mile summer road trip in extreme heat needed 1.4 more charging stops and took 68 minutes longer than the same trip in mild conditions. EV Cable Hub's 2026 US testing recorded a mean total charging time of 2 hours 14 minutes on a 500-mile trip above 100°F against 1 hour 6 minutes in mild weather.

Three effects compound on a hot road trip and none of them is large on its own. Legs get shorter, because the vehicle is losing a fifth to a quarter of its range: mean leg length fell from 152 miles to 118. Charging gets slower, because the pack arrives hot and the cabinet is derating: mean charging rate achieved fell from 108 kW to 74 kW. And each stop delivers less usable charge, because drivers left at 71% mean state of charge in heat against 78% in mild conditions, since the taper above two thirds becomes punishing on a hot pack. Two point one stops became 3.5, and 66 minutes of charging became 134.

The practical advice this produces runs against the usual guidance. In mild conditions the standard counsel is to arrive with a comfortable buffer. In extreme heat arriving lower is better, because a pack at a low state of charge sits on the fastest part of the charging curve and spends less time in the slow taper where a hot pack is penalised most. Mean arrival state of charge in the 2026 testing was 14% above 100°F against 17% in mild conditions, and the drivers who arrived lowest recorded the shortest total stop times. This trades against the safety margin a hot-weather driver wants, which is why it is offered as a finding rather than as a recommendation.

Two hidden costs sit inside each stop. 88.6% of drivers ran the air conditioning while charging, consuming a mean of 1.84 kWh per stop, or 88 cents at a 48-cent rate. And 24.1% queued for a charger in extreme heat, at a mean of 17 minutes when queuing occurred.

The failure mode that hurts most is the one the data quantifies least comfortably: 14.8% of trips found a charger derated or out of service in the heat, and the mean detour when that happened was 21 miles. In a hot climate a 21-mile detour with a hot pack and a low state of charge is a materially different event from the same detour in April. It is also why 61.4% of drivers reported at least one charging stop that was slower than they expected: the expectation was formed by a trip planner that does not model pack temperature.

• 500-mile trip, mild conditions, charging stops: 2.1
• 500-mile trip, above 100°F, charging stops: 3.5
• Extra stops: 1.4
• Total charging time, mild: 1h 06m
• Total charging time, above 100°F: 2h 14m
• Extra time: 68 minutes
• Mean leg length, mild: 152 miles
• Mean leg length, above 100°F: 118 miles
• Mean arrival state of charge, mild: 17%
• Mean arrival state of charge, above 100°F: 14%
• Mean departure state of charge, mild: 78%
• Mean departure state of charge, above 100°F: 71%
• Mean charging rate achieved, mild: 108 kW
• Mean charging rate achieved, above 100°F: 74 kW
• Charging stops that were slower than the driver expected: 61.4%
• Trips where the driver ran the air conditioning while charging: 88.6%
• Extra energy used running air conditioning while charging: 1.84 kWh per stop
• Cost of that air conditioning at 48 cents: $0.88 per stop
• Drivers who queued for a charger in extreme heat: 24.1%
• Mean queue time when queuing occurred: 17 minutes
• Trips where a charger was found derated or out of service in heat: 14.8%
• Mean detour distance when that happened: 21 miles

Summer myths tested#

Six common hot weather EV claims were tested directly against the 2026 dataset and four of them failed. Running the air conditioning did not cost more range than opening the windows at any speed above 45 mph in 2026 testing.

Each verdict below carries the figure it rests on, so the line can be lifted without the argument. Where a claim is partly true, it is marked partly true rather than forced into a verdict, because the partial cases are the interesting ones and flattening them is how myths get made in the first place.

The heat-against-cold comparison depends entirely on which question is being asked. Heat cost 21.4% of range above 95°F against 29.6% below 23°F, so cold is worse for range; heat cost 4.1 percentage points of four-year capacity while cold showed 99.4% recovery, so heat is worse for the battery.

And the overheating claim is worth correcting carefully, because it is the one that frightens prospective buyers most. Sustained power derating affected 8.6% of highway journeys above 100°F and 0.4% of stop-and-go journeys. Stop-and-go traffic at 105°F cost 24.8% of range, which is high, but it is a consumption effect rather than a thermal one, because low speed means low motor load and a car crawling in traffic is asking very little of its cooling system. The image of electric vehicles shutting down in a traffic jam is close to the opposite of what EV Cable Hub's 2026 US hot weather study measured.

• Windows down beats air conditioning: false above 45 mph. At 55 mph the drag penalty cost 4.8 miles an hour against 8.8 miles for air conditioning at 95°F, and above 45 mph air conditioning wins on comfort-adjusted value.
• Heat is worse than cold for EV range: false for range, true for battery life. Heat cost 21.4% above 95°F against 29.6% below 23°F, but heat cost 4.1 points of four-year capacity.
• Precooling wastes energy: false when plugged in. It saved 6.2 percentage points at zero cost to the pack.
• You should charge to 100% before a hot day: false. Vehicles held above 80% in heat retained 4.4 points less capacity over four years.
• DC fast charging is fine in heat: false. Power fell 31.6% above 100°F and charger-side derating affected 44.8% of cabinets above 104°F.
• Parking in the shade barely matters: false. Shade cut cabin heat soak from 138°F to 104°F and cut range loss by 3.3 percentage points.
• EVs overheat and shut down in traffic: largely false. Sustained derate affected 8.6% of highway journeys above 100°F and 0.4% of stop-and-go journeys, because low speed means low motor load.
• A hot cable is dangerous: false but relevant. Peak measured jacket temperature was 168°F, within the rating of every cable tested, but conductor resistance rose 18.4%, which is a real power loss.
• Stop-and-go traffic range loss at 105°F: 24.8%
• Stop-and-go derate incidence at 105°F: 0.4%
• Highway derate incidence at 105°F: 8.6%
• Cabin temperature after 60 minutes, sun against shade at 95°F: 138°F against 104°F
• Range loss, sun-parked against shade-parked: 21.4% against 18.1%

What US drivers actually do in the heat#

78.4% of US EV drivers parked in shade whenever it was available in 2026 and 61.2% used a windshield sunshade regularly, but only 22.1% knew their vehicle could precool the battery separately from the cabin.

The behaviour data divides cleanly into things drivers already do well and things they do not know exist. Parking and shading are close to universal among drivers who have the option: 78.4% park in shade whenever it is available, 61.2% use a windshield sunshade, and among the 42.6% who have a garage, 71.1% put the electric vehicle in it. Those are high compliance rates for behaviours that cost nothing and are visibly rewarded on the first hot afternoon.

Precooling sits in the middle. 46.8% precool on most hot days and a further 28.4% do it occasionally, which is respectable. But 34.6% of those who precool do it unplugged, where the 2026 testing shows it recovers only 1.4 percentage points instead of 6.2. Roughly a third of the precooling happening in the United States is therefore delivering a fifth of the value it could. And pack precooling, the single most valuable thing in this study per unit of effort, is known to 22.1% and used before a charging stop by 11.4%.

Charging time shifts are the clearest evidence that drivers respond to experience rather than to advice. 54.2% shifted charging to overnight because of heat, but the split by climate is stark: 71.4% in hot states against 31.8% in mild ones, and EV Cable Hub's 2026 US owner survey found the gap persists even where electricity is priced identically across the day. Hot-state drivers are not shifting to save money. They are shifting because the car charges better at night, which means they have worked out from their own sessions what the time-of-day table on this page measures.

Risk behaviour moves in summer too. 34.1% raised their minimum state of charge, and the mean minimum went from 18% in mild months to 26% in hot ones. 28.6% avoided DC fast charging entirely on the hottest days and 19.4% reduced highway speed in extreme heat. Meanwhile 38.4% report the cable is too hot to handle on hot days and 38.4% now store it indoors rather than in the vehicle: the same figure twice, which is a coincidence rather than a relationship, but a suggestive one.

The attitude figures are the honest counterweight. 87.4% would buy an electric vehicle again despite summer heat and only 7.1% cite heat as a reason they would hesitate to recommend one. But 64.2% say they were not warned about hot weather range loss at the point of purchase, and that is the number worth reporting: the problem in the United States is not that heat makes electric vehicles unsuitable, it is that nobody tells buyers what to expect before the first August.

• Park in shade whenever available: 78.4%
• Use a windshield sunshade regularly: 61.2%
• Have a garage available: 42.6%
• Park the EV in the garage where available: 71.1%
• Precool on most hot days: 46.8%
• Know about separate pack precooling: 22.1%
• Shifted charging to overnight because of heat: 54.2%
• Shifted charging to overnight in hot states: 71.4%
• Shifted charging to overnight in mild states: 31.8%
• Avoid DC fast charging on the hottest days: 28.6%
• Reduced highway speed in extreme heat: 19.4%
• Increased their minimum state of charge in summer: 34.1%
• Mean minimum state of charge, mild months: 18%
• Mean minimum state of charge, hot months: 26%
• Bought a sunshade or window tint because of EV range: 24.8%
• Had window tint applied specifically for the EV: 14.6%
• Store the charging cable indoors rather than in the vehicle: 38.4%
• Report the cable is too hot to handle on hot days: 38.4%
• Would buy an EV again despite summer heat: 87.4%
• Report heat as a reason they would hesitate to recommend an EV: 7.1%
• Say they were not warned about hot weather range loss at purchase: 64.2%

The 2026 hot weather readiness checklist#

Drivers who completed all twelve hot weather readiness actions lost 14.1% of range above 95°F (35°C) against 24.6% for drivers who completed none. That 10.5 percentage point gap is larger than the difference between most vehicles in the study.

The twelve actions are set out below in order of measured value, each with the figure attached, and they are built into an interactive checklist further down the page. Ranking them by measured value rather than by how often they are recommended produces a different order from the usual advice, and two of the top three cost nothing at all.

The gap between doing everything and doing nothing is the headline, and its size is what makes it worth publishing: 10.5 percentage points is wider than the spread between a Tesla Model 3 and a Ford F-150 Lightning above 95°F. A driver's habits are worth more than their choice of vehicle across most of the middle of the fleet. That is not true in the tails, where no habit closes the gap to a passively cooled pack, but it is true for the great majority of the vehicles in this study.

The distribution is where the opportunity sits. Only 3.4% of drivers completed all twelve actions and 21.6% completed eight or more, while 58.4% completed four or more and 9.8% completed none. Mean loss ran 14.1% for the full twelve, 16.4% for eight or more, 20.1% for four or more and 24.6% for none. The returns are not linear: the first four actions deliver most of the benefit, which means the practical advice is to pick the top four rather than to attempt the list.

In money terms the full set was worth a mean of $164 a year against completing none, most of the $214 the average driver pays for summer heat, and it excludes the four-year capacity benefit entirely.

• 1. Precool while plugged in saves 6.2 percentage points
• 2. Park in shade or a garage is worth 3.3 points of range, and 3.5 points of four-year capacity retention
• 3. Use a windshield sunshade saves 2.7 percentage points
• 4. Raise the cabin setpoint from 68°F to 74°F saves 3.4 percentage points
• 5. Use recirculation rather than fresh air saves 4.0 percentage points at 95°F
• 6. Charge overnight rather than in the afternoon is worth 10.5 points of charging efficiency and $0.92 a session
• 7. Precool the pack before a DC fast charging stop saves 12 minutes per stop
• 8. Hold state of charge between 50% and 70% in heat is worth 4.4 points of four-year capacity
• 9. Store the charging cable indoors or in shade cuts jacket temperature from 168°F to 92°F
• 10. Reduce highway speed from 75 mph to 65 mph in extreme heat saves 4.1 percentage points
• 11. Use ventilated seats to allow a higher setpoint saves 1.9 percentage points
• 12. Check tyre pressures against the hot-afternoon reading rather than the cold-morning one is worth 0.8 percentage points
• Drivers completing all twelve: 3.4%
• Drivers completing eight or more: 21.6%
• Drivers completing four or more: 58.4%
• Drivers completing none: 9.8%
• Mean range loss above 95°F, all twelve completed: 14.1%
• Mean loss, eight or more: 16.4%
• Mean loss, four or more: 20.1%
• Mean loss, none completed: 24.6%
• Mean annual saving, all twelve completed against none: $164

Apartment dwellers and street parking in extreme heat#

US EV drivers without home charging lost 24.6% of range above 95°F (35°C) against 19.8% for drivers with a garage, a gap of 4.8 percentage points. EV Cable Hub's 2026 data shows the gap is driven entirely by where the vehicle sits between journeys rather than by how it is driven.

Four effects compound for a driver in an uncovered apartment lot and each one makes the next worse. The vehicle sits in the sun, so the pack starts the day at a mean of 112°F rather than the 88°F a garaged vehicle starts at. It therefore arrives at a public charger already hot, and pulls 121 kW instead of the 152 kW a vehicle arriving from a garage pulls. That is 20.4% less power for the same stop. It pays public rates for every mile because there is no home alternative. And over four years it retains 87.1% of capacity against 90.6% for the garaged vehicle.

The cost consequence is the largest single number in this section and it dwarfs everything else on the page. Mean cost per mile was 4.82 cents for a driver with a home charger in a garage and 16.55 cents for a driver charging publicly with no home access. At 12,000 miles a year that is $578 against $1,986, a difference of $1,408 for the same miles in the same car. The summer heat penalty specifically was $412 for a driver with no home charging against the $214 study mean, so heat roughly doubles for the group least able to absorb it.

The supply side explains why the situation persists. Only 34.2% of apartment complexes in hot states have any covered parking at all, 18.6% have electric vehicle charging of any kind, and 6.1% have covered charging specifically. A resident who wants both shade and a plug has a one-in-sixteen chance of the building providing them. Mean wait for a public charger in hot states at peak evening hours was 14 minutes, spent in a vehicle cooling itself from its own pack.

Street parking is marginally worse than an uncovered lot on every measure (24.9% range loss against 24.6%, and a 114°F pack at first drive against 112°F), but the difference is small enough to say plainly that it is the absence of cover rather than the specific arrangement that does the damage. Covered parking with no charger recovered most of the thermal penalty at 21.1% loss, which is a useful finding for building owners weighing a shade structure against a charging installation: the shade is cheaper and delivers a substantial part of the benefit.

This is reported as a measurement rather than as a campaign. The 23.6% of US drivers with no home charging are the group for whom every figure on this page runs worst, and they are under-represented in EV Cable Hub's 2026 US owner survey.

• Share of US EV drivers with home charging: 76.4%
• Share without home charging: 23.6%
• Range loss above 95°F, garage parking: 19.8%
• Range loss above 95°F, covered lot: 21.1%
• Range loss above 95°F, uncovered lot: 24.6%
• Range loss above 95°F, street parking: 24.9%
• Gap, garage against uncovered lot: 4.8 percentage points
• Mean pack temperature at first drive, garage: 88°F
• Mean pack temperature at first drive, uncovered lot: 112°F
• Mean pack temperature at first drive, street: 114°F
• DC fast charging peak power on arrival from a garage: 152 kW
• DC fast charging peak power on arrival from an uncovered lot: 121 kW
• Difference: 20.4%
• Mean cost per mile, home charger in a garage: 4.82 cents
• Mean cost per mile, public charging without home access: 16.55 cents
• Annual energy cost at 12,000 miles, garage: $578
• Annual energy cost at 12,000 miles, no home charging: $1,986
• Four-year capacity retention, garaged in hot states: 90.6%
• Four-year capacity retention, uncovered in hot states: 87.1%
• Share of apartment complexes in hot states with any covered parking: 34.2%
• Share with EV charging of any kind: 18.6%
• Share with covered EV charging specifically: 6.1%
• Mean wait for a public charger in hot states at peak evening hours: 14 minutes
• Mean summer cost penalty for a driver with no home charging: $412

Tyres, pressure and hot pavement#

Tyre pressure rose 4.6 psi between a 68°F morning and a 108°F afternoon in 2026 measurement, and hot pavement raised rolling resistance by 2.1%. The net effect of hot surfaces on range was a loss of 1.3%, small enough to be the least important variable on this page and large enough to be worth stating.

Two opposing effects are at work and almost every discussion of this subject picks one and ignores the other. Hot air inside the tyre raises pressure, and a higher pressure lowers rolling resistance: the 4.6 psi rise from 38.0 psi on a 68°F morning to 42.6 psi on a 108°F afternoon was worth 0.8% of efficiency on its own. Hot pavement softens the contact patch and raises rolling resistance the other way: 2.1% on 141°F asphalt and 3.4% on 162°F asphalt, against 1.6% on 138°F concrete. Net, across the 2026 measurements, the range effect was a loss of 1.3%, rising to 2.4% at 110°F ambient on dark asphalt.

Publishing a small finding as small is the point of this section. Nothing here should change a driving decision; it exists because it is the question every enthusiast asks and nobody answers with a number.

The one genuinely actionable finding is about when pressures are checked, not about heat itself. 41.2% of drivers set their pressures from a hot afternoon reading, which leaves the tyre under-inflated when it cools overnight and costs 1.4% of range on the following cool morning. Checking against the cold-morning reading was worth 0.8 percentage points across the study. It is the twelfth and smallest item on the readiness checklist for exactly that reason, and it is included because it costs nothing at all.

Under-inflation is a much larger problem than heat is. 24.1% of drivers had at least one tyre more than 4 psi below the placard figure, and that under-inflation cost 2.4% of range, nearly twice the net effect of hot surfaces. Only 38.6% checked their pressures at any point during the summer of 2026. Tyre choice matters more still: a low rolling resistance tyre was worth 3.6% in the heat, and an all-season tyre gave up 1.9% against a summer tyre. Sustained heat also raised the tyre wear rate by 18%, which is a running cost rather than a range one but belongs in the same conversation.

• Tyre pressure at 68°F morning: 38.0 psi
• Tyre pressure at 108°F afternoon: 42.6 psi
• Rise: 4.6 psi
• Efficiency gain from that pressure rise alone: 0.8%
• Rolling resistance increase on 141°F asphalt: 2.1%
• Rolling resistance increase on 162°F asphalt: 3.4%
• Rolling resistance on 138°F concrete: 1.6%
• Net effect of hot surfaces on range: 1.3% worse
• Net effect at 110°F ambient on dark asphalt: 2.4% worse
• Tyre wear rate increase in sustained heat: 18%
• Share of drivers who set pressures from a hot afternoon reading: 41.2%
• Range cost of setting pressures hot then driving in the cool morning: 1.4%
• Measured advantage of a cold morning reading: 0.8 percentage points
• All-season tyre efficiency against summer tyre in heat: 1.9% worse
• Low rolling resistance tyre advantage in heat: 3.6% better
• Tyre temperature after 60 minutes at 75 mph on 141°F asphalt: 168°F
• Tyre pressure at that temperature: 45.1 psi
• Share of drivers who checked pressures during summer 2026: 38.6%
• Share who had at least one tyre more than 4 psi below placard: 24.1%
• Range cost of that under-inflation: 2.4%

Heat, grid outages and using the vehicle as a power source#

18.4% of US EV drivers in hot states experienced a grid outage during a heat event in 2026, and 11.2% used their vehicle to power something in the home during one. A 77 kWh pack at 80% charge ran a 900 W window air conditioning unit for 54 hours in EV Cable Hub's 2026 testing.

Heat events and grid stress arrive together, which makes vehicle-to-load a genuinely useful hot weather capability rather than a novelty feature. The mean outage during 2026 heat events ran 4 hours 20 minutes and the longest recorded in the survey ran 38 hours. An electric vehicle sitting in the driveway with 60 usable kilowatt hours in it is, in that situation, one of the largest energy stores most households own.

The runtime figures are what make the capability concrete, and they are calculated from a 77 kWh pack at 80% state of charge down to the 20% cutoff at which most vehicles stop supplying. A 900 W window air conditioning unit ran 54 hours. A 1,500 W portable unit ran 32 hours. A 150 W refrigerator ran 324 hours and a 200 W chest freezer 243 hours. A 40 W CPAP machine ran 1,215 hours, which is more than seven weeks. That is the single most reassuring number in this section for anyone who depends on medical equipment through a summer outage.

Two practical constraints matter more than the headline capability. Only 34.6% of vehicles in the survey support vehicle-to-load at all, so the first thing to check is whether the car can do it before planning around it. And measured output ran below published rating: a mean of 2.94 kW against a mean published rating of 3.2 kW, a shortfall of 8.1%. That is enough to matter when sizing against a 1,500 W air conditioner plus a refrigerator, and it is not disclosed anywhere at the point of purchase.

The energy cost is worth stating in the units drivers think in. Running a 900 W unit for 24 hours consumes 21.6 kWh, which on a 3.12 mi/kWh vehicle is 67 miles of range. That is a straightforward trade in a heat emergency and an expensive one otherwise, and it is why the state-of-charge behaviour in the survey makes sense: 34.1% of drivers held a higher state of charge during heat warnings, at a mean of 78% against 62% on an ordinary summer day, and 28.4% charged pre-emptively before a forecast heat event.

That pre-emptive charging sits in direct tension with the degradation finding earlier on this page, where holding above 80% in heat cost 4.4 percentage points of four-year capacity. Both are correct. Holding 78% for the two or three days a heat warning is in force is a sensible insurance premium; holding it all summer is not. Readers with a capable vehicle will find the equipment side in our vehicle-to-load adapter range.

• Share of hot-state drivers experiencing an outage during a heat event: 18.4%
• Mean outage duration during 2026 heat events: 4h 20m
• Longest outage recorded in the survey: 38 hours
• Share who used their vehicle to power something: 11.2%
• Share whose vehicle supports vehicle-to-load: 34.6%
• Mean published vehicle-to-load rating across capable vehicles: 3.2 kW
• Mean measured output: 2.94 kW
• Mean shortfall against rating: 8.1%
• Runtime for a 900 W window air conditioner from a 77 kWh pack at 80%: 54 hours
• Runtime for a 1,500 W portable air conditioner: 32 hours
• Runtime for a 60 W ceiling fan: 810 hours
• Runtime for a 150 W refrigerator: 324 hours
• Runtime for a 200 W chest freezer: 243 hours
• Runtime for a 40 W CPAP machine: 1,215 hours
• Runtime for a 1,200 W microwave in intermittent use: 41 hours
• Mean minimum state of charge cutoff before the vehicle stops supplying: 20%
• Share of drivers who kept a higher state of charge during heat warnings: 34.1%
• Mean state of charge held during a heat warning: 78%
• Mean state of charge held on an ordinary summer day: 62%
• Share who charged pre-emptively before a forecast heat event: 28.4%
• Energy cost of running a 900 W unit for 24 hours from the pack: 21.6 kWh
• Range cost of that energy on a 3.12 mi/kWh vehicle: 67 miles

Interactive tools#

Three calculators built on the 2026 dataset, a state comparator covering all 44 states, a searchable table of every figure on this page, and the twelve-item readiness checklist. Everything runs in the browser.

Each tool draws on the tables above rather than on a separate dataset, so every figure behind them is published on this page and can be checked against it. Where a tool applies an adjustment, the value it replaces is shown in the option label so the arithmetic is visible.

Methodology#

Every figure on this page comes from one of four EV Cable Hub studies conducted between April and July 2026: a range study of 18,640 US summer journeys, a heat charging test of 3,180 monitored sessions, a bench programme covering 68 charging cables, and a survey of 2,310 US drivers.

Frequently asked questions#

Thirty questions on hot weather EV range loss, each answered with the 2026 figure first.

Every answer below is drawn from the tables on this page. Temperatures are given in Fahrenheit with Celsius alongside, and where a figure is an association rather than a measured causal effect it is described as such.