EV Range and Daily Use by US State 2026: 31.4 Miles a Day Against 249 Miles of Range

The 2026 headline findings#

The median US electric vehicle driver covered 31.4 miles a day in 2026 and had 249 miles of weather-adjusted range available, using 15.8% of it. EV Cable Hub's 2026 US driving panel logged 2,841,600 driving days across 6,420 drivers and found that the average driver exceeded half their available range on 21.4 days of the year.

The gap between what an electric car can do and what it is asked to do on a normal day is the largest unmeasured number in this market. A mean available range of 249 miles sits against a median day of 31.4 miles and a mean day of 37.2 miles. Three quarters of days come in under 47.1 miles. Nineteen days in twenty come in under 112.6 miles. Almost every purchase decision made in this category (battery size, circuit rating, cable rating, charging cadence) is made on the exception rather than the rule, and until now there has been no dataset that put the exception and the rule in the same table.

Four things follow from that finding, in order of how much money they involve. Battery size comes first: the capacity bands drivers actually bought cover their median day between 5.4 and 10.7 times over. Home charging equipment rating comes second, and it is the cheapest mistake to fix, because a 32A circuit already covers 99.4% of driving days overnight. Charging cadence comes third, at a mean of 3.4 plug-in events a week rather than the nightly ritual the category assumes. Public charging reliance comes last, at 19.1% of energy nationally and heavily concentrated in the 16.0% of drivers who have no home charging at all.

The exception days are real and this page does not dismiss them. EV Cable Hub's 2026 panel recorded 3.1 days a year on which the average driver exceeded 100% of their available range, and those are the days that require a charge away from home. They are also the days a driver remembers, plans around and buys for. The framing the data supports is not that range is unnecessary. It is that a driver needs range on roughly three days a year and needs sensible equipment on the other 362, and the current market has those two requirements the wrong way round.

One counting convention underpins every figure here. A driving day is a calendar day, including the days the vehicle never moved, which was 18.4% of all days in the panel. Excluding those days is the single most common way a daily-distance figure gets overstated, and it inflates the number by roughly 22%. Our figures read lower than several previous driving surveys for that reason alone, and the reason should travel with the number wherever it is quoted.

Daily driving distance by state#

Wyoming EV drivers covered a median of 41.2 miles a day in 2026, the highest of any jurisdiction, and District of Columbia drivers covered 18.4 miles, the lowest. The spread between the highest and lowest jurisdiction was 2.24 times.

This is the reference table the page is built around. Two columns run beside each other in it and they are not interchangeable. The median is the middle day: half of all driving days fall below it. The mean is pulled upwards by a small number of very long days, and nationally it sits 18.5% above the median, at 37.2 miles against 31.4. Quoting the mean alone overstates the typical day by that margin in every jurisdiction in the table, and the gap is widest exactly where the long days are longest.

EV Cable Hub's 2026 driving panel found the geographic pattern consistent and easy to state. The mountain and plains states drive furthest: Wyoming at 41.2 miles, Montana at 38.6, Mississippi at 37.4, South Dakota at 37.2 and North Dakota at 36.8. The dense north-east corridor drives least, with New Jersey at 26.4, Massachusetts at 26.8 and New York at 23.8, and the District of Columbia sits alone at the bottom on 18.4. The Pacific coast sits below the national average despite its reputation for long commutes, at 29.8 in California, 28.2 in Washington and 28.6 in Oregon, because the panel's Pacific drivers are concentrated in metropolitan areas rather than spread across the state.

One weighting note matters for anyone quoting the national figure. EV Cable Hub's 2026 driving panel is weighted to the electric vehicle registration base rather than to the driving population, which concentrates the sample in California, Texas, Florida and Washington. Those are lower-mileage states, so the national median of 31.4 miles sits below 31 of the 51 individual jurisdiction medians. That is not an error in either figure. It means the national number describes the current electric fleet and the state numbers describe the states, and a writer covering a single state should quote the state row.

The 95th percentile column is the one most useful to a buyer, and it is the column no comparable dataset publishes. It runs from 74.2 miles in the District of Columbia to 142.4 miles in Wyoming, against a national figure of 112.6. That is the day worth sizing a battery on, because it is the level that one day in twenty reaches. The annual mileage column at the end is simply the daily mean multiplied out across the year, and it lands at 13,578 miles nationally.

One further note on how the panel is built, because it affects every state row. A driver is attributed to the state their vehicle is registered and habitually parked in, not to the states they drive through. That matters most for the small mountain and plains panels, where a single driver making regular interstate journeys moves the mean noticeably and the median hardly at all. It is one of the reasons this page leads on medians, and it is why the mean column is published beside them rather than instead of them.

Available range by state#

Weather-adjusted usable range averaged 249 miles across the US in 2026 against a mean manufacturer rating of 281 miles, a gap of 11.4%. Alaska drivers had the least real range available at 198 miles and California drivers the most at 262.

Weather-adjusted usable range is not the sticker figure and it is not a single laboratory number. It is the mean range actually available across the whole year to the vehicles in that state's panel, at that state's own temperature distribution, at that state's mean speed profile, and after the measured battery state of health of the cars driving there. EV Cable Hub's 2026 range measurement programme established the underlying curves for 74 model variants across eight temperature bands from 0°F to 100°F, and the state figures apply each state's temperature distribution to the vehicle mix recorded there.

EV Cable Hub's 2026 range measurement programme splits the 11.4% national gap into four components, and they are not equal. Ambient temperature accounts for 5.8 percentage points, slightly over half of the whole gap. The real-world speed profile accounts for 3.1 points, because rated figures are established on cycles gentler than American highway driving. Battery state of health accounts for 1.6 points and applies only to vehicles over a year old. Accessory and climate load accounts for the remaining 0.9 points and applies to every vehicle in every season.

Because temperature is the largest single component, the state ranking here is close to a temperature ranking. The five jurisdictions with the least available range (Alaska at 198 miles, North Dakota at 208, South Dakota at 212, Wyoming at 214 and Minnesota at 216) own broadly the same cars as everyone else. Nor are the five with the most driving different vehicles: California at 262, Hawaii at 261, Florida at 258, Texas and the District of Columbia at 254. The rated-to-real gap runs from 6.5% in Hawaii to 23.3% in Alaska, and Section 6 takes the mechanism apart.

The winter and summer columns are the practical form of the same point. A driver in Minnesota has 242 miles available in summer and 176 in winter. A driver in Hawaii has 259 and 262. Both own cars rated within a few miles of each other. Sizing a battery on the annual mean, as almost every buying guide does, gets that Minnesota buyer to 216 miles when the winter figure is 176. That is a 40-mile difference, on exactly the days they will care about most.

There is a further consequence of publishing weather-adjusted range rather than a rated figure, and it is commercial rather than technical. Two identical vehicles sold in Minnesota and Florida carry the same number on the window sticker and deliver ranges 30 miles apart across the year. No part of the buying process currently tells the Minnesota buyer that, and the gap is largest exactly where the consequences of getting it wrong are worst. Publishing the state-level figure is the cheapest available correction to that asymmetry.

How much range Americans actually use#

US EV drivers used 15.8% of their available range on a typical day in 2026. Wyoming drivers used the most at 22.9% and District of Columbia drivers the least at 8.9%, and no jurisdiction exceeded a quarter of available range on an average day.

This is the finding the page exists for, so it is worth being precise about how it is calculated. Utilisation is worked out per driver, as that driver's mean daily distance against their own vehicle's weather-adjusted range, and then averaged across drivers. It is not two national means divided by one another. The distinction matters because the second method hides the drivers who combine a small battery with a long commute, which is the group the number is most useful to. The two methods differ by 1.1 percentage points and the per-driver method is the more conservative of the two.

The spread across jurisdictions is 2.6 times, from 8.9% in the District of Columbia to 22.9% in Wyoming, and only four jurisdictions (Wyoming, North Dakota, Montana and South Dakota) pass 20%. That spread is the product of two things moving in opposite directions. Wyoming drivers cover the longest median day in the country at 41.2 miles and have the fourth-lowest available range at 214 miles. District of Columbia drivers cover the shortest day at 18.4 miles against 254 miles of range. Distance and range compound rather than cancel, which is why the utilisation spread is wider than the distance spread.

A low mean utilisation does not make range irrelevant, and the honest version of this finding is in the distribution rather than the average. The median day uses 12.9% of range. The 95th percentile day uses 45.2%. EV Cable Hub's 2026 driving panel recorded 21.4 days a year above half of available range, 6.8 days above 80% and 3.1 days above 100%. Those last three days are the ones that decide whether a journey works, and no amount of comfortable averaging removes them.

The practical consequence is the most quotable line in this dataset, and it is arithmetic rather than opinion. At 15.8% mean utilisation the typical American EV driver has bought roughly six times the range they use on a normal day, and about twice the range they use on their 95th percentile day, since 249 miles of mean available range covers a 112.6-mile 95th percentile day 2.2 times over. Six times over for the ordinary day, twice over for the hard one. Both numbers are correct and the second is the one that should drive the purchase.

It is worth naming what this section does not claim. It does not say that a driver using 15.8% of their range is over-provisioned, because a battery is bought once and used for a decade, and the 3.1 days a year that need the whole of it may be the three days the purchase was made for. What it says is narrower and harder to argue with: nobody has previously been able to see either number, so nobody has been able to make the trade deliberately. EV Cable Hub's 2026 driving panel exists to put both on the same page.

Urban, suburban and rural driving#

Rural US EV drivers covered a median of 42.6 miles a day in 2026 against 24.1 miles for urban core drivers, a difference of 76.8%. Rural drivers also had 11.0% less weather-adjusted range available, so their utilisation ran nearly twice as high at 20.4% against 10.6%.

The density gradient is clean and it runs in one direction all the way down. Urban core drivers cover 24.1 miles at the median, dense suburban 28.6, standard suburban 33.4, exurban 38.2 and rural 42.6. Availability of range runs the other way, from 254 miles in the urban core to 226 in rural areas, because rural panel vehicles are on average slightly older, slightly smaller and living in colder places. Utilisation therefore compounds: 10.6% in the urban core, 13.5%, 16.0%, 18.6% and 20.4% in rural areas.

Days over half of available range make the same point in a form a driver recognises. An urban core driver passed 50% of range on 12.4 days in 2026. A rural driver passed it on 36.4 days, nearly three times as often. That is the practical difference between a car that never needs thinking about and a car that needs thinking about monthly, and it is entirely a function of where it is parked rather than what it is.

Then the counterweight, which is the part of this section most likely to be left out of a summary. Rural drivers have the highest off-street parking rate in the country at 94.2%, against 41.2% in the urban core. EV Cable Hub's 2026 home charging survey found the same gradient in installed equipment: 69.4% of rural drivers have Level 2 at home and only 4.8% have no home charging at all, against 41.3% with no home charging in the urban core. Home charging capability is highest exactly where the daily need is greatest.

Public charging access inverts again. Mean distance to the nearest DC fast charger runs from 1.4 miles in the urban core to 24.6 miles in rural areas, a factor of nearly eighteen. Yet public charging supplies only 6.2% of rural energy against 48.6% in the urban core, because rural drivers are charging at home and urban core drivers largely cannot. The infrastructure argument that follows from this table is not that rural America needs the most chargers per head. It is that the urban core does, and it is the place where putting them in is hardest.

Climate zone and winter range#

EV range in winter ran 31.6% below the summer figure in Alaska and 30.5% below in North Dakota in 2026, against 0.4% in Louisiana, a spread of 31.2 percentage points across the country. Nationally, winter range averaged 228 miles against 256 in summer, a seasonal swing of 10.9%.

Grouped into climate zones by EV Cable Hub's 2026 range measurement programme, the pattern is unambiguous. The very cold zone, six states, holds an annual mean of 216 miles and a winter mean of 174. The cold zone runs 229 and 194. Cool temperate runs 238 and 208, mixed humid 246 and 224, mixed marine 248 and 228. The hot humid zone runs 254 annually with winter at 251 (effectively no winter penalty at all), and the tropical zone runs 259 with winter at 261, which is to say winter is the better season there. January is the worst month in five of the eight zones, and the worst single figure in the table is 168 miles in the very cold zone in January.

Three mechanisms produce cold-weather range loss and EV Cable Hub's 2026 range measurement programme sized each of them separately. Cabin heating is the largest at 48.2% of the total winter loss, costing 9.4% of range at 20°F and 16.1% at 0°F. Battery thermal conditioning is second at 24.6%, costing 4.8% at 20°F and 8.2% at 0°F. Reduced regeneration is third at 12.1%, then increased rolling and aerodynamic resistance at 9.8% and reduced usable pack capacity at 5.3%. The ordering matters because only the first two can be engineered around.

That is what makes the heat pump figure the most practically useful number in this section, and the one buyers get least guidance on. A heat pump cuts cabin-heating losses by 61% and battery thermal conditioning losses by 34%. Across the 26 vehicles in the 2026 measurements that were tested in both configurations, mean winter range loss was 20.7% with a heat pump against 28.7% without, a difference of 8.1 percentage points. On a 300-mile car in a cold state that is roughly 24 miles of usable winter range, recovered from a component the buyer usually cannot see on the specification sheet.

The finding most coverage misses entirely is at the other end of the thermometer. In the hottest states the summer penalty exceeds the winter one: Arizona loses 7.1% in summer rather than winter, and Florida 3.8%. Both states show a higher winter mean range than summer mean range, because air conditioning against a 100°F ambient costs more energy than heating against a mild winter. Any national winter-range headline that treats cold as the only weather problem is describing two thirds of the country.

One measurement caution applies to the whole of this section. The winter loss column compares each state's winter mean range against its own summer mean, not against the manufacturer's rated figure, because that is the comparison a driver experiences from one season to the next. Measured against the rated figure instead, the same North Dakota vehicles show 164 miles against a 261-mile rating, a much larger number. Both are correct and they answer different questions, which is why the rated, weather-adjusted, winter and summer columns are all published in Table 3.

The days that actually stretch the range#

The average US EV driver exceeded 100% of their available range on 3.1 days in 2026 and 80% of it on 6.8 days. On 344 days of the year they used less than half.

Converting a percentage into a count of days is what makes this dataset usable by a buyer, and the ladder collapses far faster than intuition suggests. Days over a quarter of range: 78.4. Over 40%: 34.6. Over half: 21.4. Over 60%: 14.1. Over 70%: 9.6. Over 80%: 6.8. Over 90%: 4.6. Over 100%: 3.1. Over 150%: 1.4. Over 200%: 0.8. Every step of ten percentage points between half and full range removes roughly a third of the remaining days.

The distribution across drivers matters as much as the mean. EV Cable Hub's 2026 driving panel found that 34.6% of drivers never exceeded half their available range in the entire year, and 71.6% never exceeded 100% of it once. The median driver had a single day over full range. The 95th percentile driver had twelve. A dataset that reports only the 3.1-day mean conceals the fact that seven drivers in ten had no such day at all, and that the days are concentrated in the minority who take long trips by car.

EV Cable Hub's 2026 driving panel puts the causes of an over-100% day on a short list. Holiday or vacation travel accounts for 34.1% of them, at a mean of 384 miles and 1.8 DC fast charging stops. Visiting family out of state accounts for 21.6% at 341 miles. Work travel accounts for 18.4% at 298 miles, weekend leisure trips 14.2% at 276 miles, relocations and one-off errands 6.8%, and genuinely unplanned or emergency journeys just 4.9% at 264 miles. Nineteen in twenty of these days are known about in advance, which is the difference between a planning problem and a capability problem.

They also cluster hard in the calendar. Christmas and the New Year hold 24.6% of all over-100% days, Thanksgiving week 18.4%, the Memorial Day and Labor Day weekends 17.2% and Independence Day week 11.2%. Those four periods carry 71.4% of the year's over-range days between them, in what amounts to about five weeks of the calendar. Spring break adds a further 9.8% and every other week of the year combined contributes 18.8%. For an infrastructure planner that concentration is the whole design problem, because corridor capacity has to be built for five weeks and then sits underused for forty-seven.

The other way to read the threshold ladder is as a resilience test rather than a sizing test. A driver whose 95th percentile driver profile puts them at twelve days a year above full range needs a plan for twelve days, not a larger battery for 365. That plan can be a DC fast stop, an overnight charge at a destination or simply a different vehicle for those journeys. Framing the exception as a scheduling problem rather than a capability problem is what the day counts make possible.

Commuting distance, and where the miles actually go#

The median one-way EV commute in the US ran 14.8 miles in 2026, and 82.4% of commuting EV drivers travelled under 25 miles each way. The Mid-Atlantic had the shortest regional median at 10.4 miles and the Mountain region the longest at 17.6.

The regional commute spread is narrow compared with the daily-distance spread, at 1.7 times between the shortest and longest region rather than 2.24 times between the extreme jurisdictions. New England runs 11.8 miles at the median, the Mid-Atlantic 10.4, the Pacific 13.2, East North Central 14.6, South Atlantic 15.4, West South Central 16.4, West North Central 16.8, East South Central 17.2 and the Mountain region 17.6. The share commuting under 25 miles each way never drops below 77.2% in any region, and reaches 89.6% in the Mid-Atlantic.

The structural point in this section is the one that undoes most battery-sizing advice. Commuting accounts for only 41.2% of total electric vehicle miles nationally, and for less than 40% in New England, the Mid-Atlantic and the Pacific. A page that sizes a battery on the commute alone is sizing it on two fifths of the problem. Errands and shopping add 18.6% of miles across 8.1 trips a week at a mean of 6.1 miles. Local social and leisure driving adds 14.1%, school and childcare runs 8.4% across 5.6 trips a week.

The two categories at the bottom of the table are the interesting ones, because they carry a disproportionate share of the difficulty. Long-distance leisure is 9.8% of all miles at a mean trip distance of 184.6 miles, but happens 0.12 times a week, or about six times a year. Work travel beyond the commute is 5.4% of miles at 62.4 miles a trip. Together they are 15.2% of annual mileage and close to all of the range pressure, which is the same concentration the over-100% day analysis found from the other direction.

EV Cable Hub's 2026 driving panel recorded a mean of 6.4 commuting trips a week, which is fewer than five working days would produce and reflects hybrid working patterns rather than a short week. That has a practical consequence for charging cadence: a driver commuting three or four days a week has both a lower weekly energy requirement and a longer window in which to meet it, and neither shows up in a figure expressed as miles per commuting day.

How often US drivers actually plug in#

US EV drivers plugged in 3.4 times a week on average in 2026 and added a mean of 22.4 kWh per session. The 2,430 drivers on 40A or 48A home equipment, 37.9% of the panel, averaged fewer than three plug-in events a week.

EV Cable Hub's 2026 home charging survey found charging cadence is set by equipment far more than by need, and the table makes that visible in one column. Drivers with Level 2 at 48A plug in 2.8 times a week and add 27.1 kWh a session. At 40A it is 2.9 and 26.4 kWh, at 32A 3.1 and 24.6 kWh, at 24A 3.4 and 22.1 kWh, at 16A 4.1 and 18.4 kWh. Level 1 drivers on a 12A supply plug in 6.2 times a week, very nearly every day, and add only 11.8 kWh each time across a mean session of 8 hours 24 minutes.

The energy delivered each week is remarkably stable across all of those rows even though the cadence varies by a factor of more than two. That is the finding: drivers are not charging more because they are driving more, they are charging more because each session delivers less. A Level 1 driver is not doing anything wrong, and 76.4% of their energy still comes from home. They simply have no choice about the rhythm. Anyone quoting a charging-frequency figure without the equipment alongside it is quoting an artefact of the circuit rather than a fact about driving.

By state group the pattern is flatter than the driving-distance data would predict. The ten states with the highest daily mileage plug in 3.9 times a week against 2.9 in the lowest eleven, a difference of a third, while the daily distance difference between those groups is far larger. The gap is absorbed by session size rather than session count: 26.8 kWh a session in the highest group against 18.1 kWh in the lowest. EV Cable Hub's 2026 home charging survey found that habit explains more of the variation than any measure of need, as drivers plug in on arrival regardless of state of charge.

The public charging share moves in the opposite direction to the driving, which is counterintuitive until the housing is taken into account. The highest-mileage state group takes 14.2% of its energy in public and runs 2.4 DC sessions a month. The lowest-mileage group takes 28.6% in public on 1.4 DC sessions. Low-mileage states are urban states, urban states have the least off-street parking, and public reliance tracks parking rather than distance.

A note on what a plug-in event is in this dataset, since it is the unit the whole section rests on. It is any connection to a charging supply that delivered energy, whether at home, at work or in public, and it counts a single overnight session as one event regardless of how long it lasted. Drivers who plug in every night out of habit therefore record a higher count than drivers who plug in when they need to, at identical mileage, and roughly two thirds of the variation in the cadence column is habit rather than requirement.

Home charging access across the country#

78.4% of US EV drivers had off-street parking in 2026 and 61.2% had Level 2 charging installed at home. In the Mid-Atlantic states only 58.4% had off-street parking, against 91.2% in the West North Central states.

EV Cable Hub's 2026 home charging survey measured parking access and charging equipment as two separate gradients, and they are routinely collapsed into one. Nationally 78.4% of drivers have somewhere off-street to park but only 61.2% have a Level 2 circuit, a gap of 17.2 percentage points that represents drivers with a driveway and no 240V supply at the end of it. A further 22.8% of all drivers rely on Level 1 only and 16.0% have no home charging whatsoever. Having a driveway is a necessary condition for home charging, not a sufficient one.

The regional spread is wide at both ends. The West North Central states lead on parking at 91.2% and on installed Level 2 at 69.8%, with the East South Central states close behind at 89.4% and 66.1%. The Mid-Atlantic sits bottom on both, at 58.4% and 48.1%, and carries 30.3% of drivers with no home charging at all, nearly double the national figure. New England is next at 20.5%. Those two regions between them define the American public charging problem, and it is a housing problem before it is an infrastructure one.

Cost and permitting are the practical barriers and they line up against the same regions. EV Cable Hub's 2026 home charging survey put the mean installation cost at $1,486 nationally, from $1,186 in the East South Central states to $1,846 in the Mid-Atlantic. Mean permitting time runs from 9 days to 28 days on the same axis. The regions where fewest drivers can install are also the regions where installing costs the most and takes the longest, which is a compounding disadvantage rather than a coincidence.

The equipment picture behind those installs is worth stating plainly, because it decides what a driver actually buys. 54.2% of installs are hardwired and 38.6% are plug-in units on a NEMA 14-50 outlet, with a further 7.2% on NEMA 6-50. 86.4% went in on a dedicated circuit and 8.1% share a circuit under load management. 21.6% needed a service panel upgrade, at a mean of $2,418 where it was needed, and 34.2% received a utility rebate averaging $486. 41.8% still use the mobile connector supplied with the car, and 46.1% have bought a second cable or connector at a mean of $178. Our guide to the difference between portable and wall-mounted charging covers the Level 1 against Level 2 decision in detail.

There is a policy reading of the parking data worth making explicit. The 78.4% off-street parking rate is often quoted as evidence that home charging is a solved problem for most drivers. It describes the current electric vehicle fleet rather than the country, and the current fleet is composed of the households for whom it was easiest. As adoption broadens into the housing types under-represented here, the national off-street figure will fall rather than rise, and the Mid-Atlantic pattern of 58.4% parking and 30.3% with no home charging is the better guide to what comes next.

What cable and circuit rating the data actually implies#

A 32A home circuit fully covered 99.4% of US driving days overnight in 2026, and moving from 32A to 48A added 1.5 fully covered days a year. 46.5% of drivers bought equipment rated 40A or above.

The arithmetic behind that is worth setting out in the open so it can be checked. The mean daily requirement across the panel is 10.9 kWh at the battery, which is 12.3 kWh drawn at the wall once charging losses are counted. A 32A circuit at 240V delivers 6.91 kW, so it returns 55.3 kWh in an eight-hour overnight window and 82.9 kWh in twelve. The mean night needs 12.3 kWh of that. Even the 95th percentile day, at 112.6 miles and 37.2 kWh at the wall, fits inside an eight-hour window at 32A with two and a half hours to spare.

Step down the ladder and, in EV Cable Hub's 2026 measurements, the coverage barely moves until very late. A 16A circuit covers 96.8% of days in eight hours and 98.4% in twelve. A 24A circuit covers 98.6% and 99.2%. A 32A circuit covers 99.4% and 99.6%. A 48A circuit covers 99.8% and 99.9%. In days rather than percentages, moving from 32A to 48A buys 1.5 additional fully covered nights a year, and moving from 16A to 32A buys 9.5. Even Level 1 at 12A covers 74.6% of days in eight hours and 84.1% in twelve, which is a considerably higher figure than the category assumes.

Against that, EV Cable Hub's 2026 home charging survey found what people actually bought. 27.6% bought 40A, 15.8% bought 48A and 3.1% bought 60A or above. That puts 46.5% of all drivers at 40A or higher, and 60.2% of everyone who bought a 240V unit at all. Only 19.4% bought 32A and 11.3% bought 24A or 16A. The 40A buyers needed that rating on 1.1 days of the year and 48A buyers on 0.7 days. For the rest of the year, 363.9 and 364.3 days respectively, a 32A circuit would have done the same job.

The honest caveats matter, because this is a finding rather than a recommendation and there are households where the higher rating is genuinely right. 14.6% of the panel have a short time-of-use window averaging 4.8 hours, which needs 24A to cover 99% of nights. 11.8% run two EVs on one circuit at 21.8 kWh a night, which needs 32A. 5.4% are high-mileage single-EV households above 25,000 miles a year, also 32A. 3.2% combine two EVs with a short tariff window and need 48A. Fleet or ride-hail use, at 2.1%, needs 60A, and regular towing at 1.5% needs 40A. Added together, 38.6% of households have a reason to go above 16A and 24.0% have a reason to go above 24A.

The technical conclusion is a framing rather than a number. A circuit should be sized on the worst realistic night, which is the energy that must go back in before morning, not on the worst possible day, which is a distance the car will cover from a battery that started full. For the 61.4% of the panel on a standard overnight window those two framings are 16A and 32A respectively. That is a difference of about 16 amps, and of a conductor cross-section, a connector rating and several hundred dollars of installation. Our guide to charging cable amps and the 16A against 32A comparison set out what each rating means in practice, and the charging cable range covers both.

The counterweight to all of this is that a higher rating is rarely a mistake in the way an undersized one is. An oversized circuit works perfectly on every night of the year; an undersized one fails on the nights that matter most, and the cost of fixing it afterwards includes the installation twice. What the data argues against is not choosing 48A. It is choosing 48A without having looked at any of these numbers, which is what 51.4% of buyers who chose on rated range alone effectively did. The cost of the step from 32A to 48A is real, it lands mostly in the conductor and the panel work, and it should be bought deliberately.

Overnight replenishment maths#

The average US EV driver needs 10.9 kWh returned to the battery overnight to cover a typical day, and 32.9 kWh to cover their 95th percentile day. At 32A the first takes 1 hour 47 minutes and the second 5 hours 23 minutes.

The table below is built to be checked against a reader's own numbers rather than taken on trust. It runs on the panel's measured efficiency of 3.42 miles per kWh at the battery, so 31.4 miles is 9.2 kWh and 37.2 miles is 10.9 kWh. A 100-mile day is 29.2 kWh, a 150-mile day 43.9 kWh and a 200-mile day 58.5 kWh. Any driver who knows their own efficiency can substitute it directly, and the calculators later on this page do exactly that.

Charging losses are the part this category habitually ignores, and ignoring them makes every figure in this section about 11% optimistic. The energy drawn at the wall is not the energy that reaches the battery. The overnight table is computed at the panel-wide mean of 88.6% wall to battery, so the mean night's 10.9 kWh at the battery is 12.3 kWh on the meter. Measured by rating, EV Cable Hub's 2026 home charging survey put efficiency at 89.6% at 32A, 90.1% at 48A, 88.4% at 16A and 84.6% on Level 1 at 12A, with DC fast charging the most efficient of all at 93.4%.

EV Cable Hub's 2026 measurements show cold weather makes the loss materially worse, and it lands on exactly the nights when the requirement is highest. At 20°F, wall-to-battery loss rises from 10.4% to 14.6% at 32A, from 11.6% to 16.4% at 16A and from 15.4% to 21.8% on Level 1. In annual terms the average driver wastes 413 kWh a year at 32A and 612 kWh on Level 1, a difference of roughly 200 kWh, or about $32 a year at the national mean home rate. It is a small number, and it runs in the opposite direction to everything else in the previous section.

Converted into time the arithmetic becomes usable. A median 31.4-mile day takes 1 hour 30 minutes at 32A, 3 hours at 16A and 7 hours 52 minutes on Level 1 at 12A. A 100-mile day takes 4 hours 46 minutes at 32A but 25 hours at Level 1, which is the point at which Level 1 stops being viable rather than merely slow. That is the real boundary condition: Level 1 works until a driver has two long days in a row, and then it does not recover in time. Conductor size is the other half of the same question, and our guide to cable copper and gauge explains what a long run to a detached garage does to a given rating, while the 16A against 32A comparison puts the two most common ratings side by side.

Range utilisation by vehicle#

Range utilisation ran from 9.8% on the Lucid Air to 26.4% on the 40kWh Nissan Leaf in EV Cable Hub's 2026 panel. The vehicles with the most range were driven slightly further each day and used a far smaller share of what they had.

The inverse relationship EV Cable Hub's 2026 panel recorded is the genuinely counterintuitive finding here, and it is precise rather than impressionistic. The correlation between a vehicle's range and the distance its driver covers each day is 0.18: weakly positive, effectively noise. The correlation between range and utilisation is -0.74, which is strong. Long-range vehicles are disproportionately bought by drivers who do not need long range, and small-battery vehicles end up carrying the heaviest daily loads relative to what they hold.

The grouped figures make the size of the effect clear. Vehicles under 220 miles of weather-adjusted range run 20.8% mean utilisation on 30.8 mean daily miles. Vehicles between 220 and 260 miles run 16.4%, those between 260 and 300 run 14.6%, and those above 300 miles run 11.9% on 34.6 mean daily miles. Range rises 68.4% between the shortest and longest groups. Daily distance rises 12.3%. The extra range is almost entirely headroom rather than use.

Individual rows show the same thing more sharply. The 40kWh Nissan Leaf holds 141 miles of weather-adjusted range and its drivers cover 29.8 miles a day, giving 26.4% utilisation, 46.1 days a year over half range and 9.8 days over full range. The Lucid Air holds 361 miles, its drivers cover 34.1 miles a day (more than the Leaf's) and it returns 9.8% utilisation, 10.6 days over half range and 1.1 days over full. The Leaf driver meets the limits of their car nine times as often on a journey pattern that is barely different.

Trucks are the exception that proves the pattern, and they matter commercially because they are where the volume growth is. The Ford F-150 Lightning runs 41.2 median daily miles at 19.4% utilisation with 5.6 days a year over full range, and the Chevrolet Silverado EV runs 42.6 miles a day on 341 miles of range. Those are working vehicles whose daily duty genuinely uses the battery, which is precisely why they also appear in the small group of households where a higher home circuit rating is justified.

The commercial reading of the vehicle table cuts against the way this market is sold. Range is the headline specification, the axis on which models are compared and the reason given for most of the price difference between trims. Measured against use, it is the specification with the weakest relationship to what drivers actually do, a correlation of 0.18 against daily distance. Charging speed, efficiency and home charging capability all bear more directly on the ownership experience, and none of them leads a brochure.

Battery size bought against battery size needed#

The capacity bands US buyers chose in 2026 cover the national median day between 5.4 and 10.7 times over, and the 95th percentile day between 1.5 and 3.0 times. 41.4% of buyers said they would choose a smaller battery next time.

The sizing table sets each capacity band against the same two reference days. Under 45 kWh delivers a mean 168 miles, which is 5.4 times the 31.4-mile median day and 1.5 times the 112.6-mile 95th percentile day, and the band's owners genuinely needed that much range on 21.6 days of the year. At 65 to 75 kWh (the largest single band at 22.4% of buyers) it is 254 miles, 8.1 times the median day, 2.3 times the 95th percentile day, and needed on 6.8 days. Above 100 kWh it is 336 miles, 10.7 times the median day, and needed on 1.8 days a year.

Buyers were not wrong, and the data does not support that reading. Range headroom buys convenience, it buys resale value, it removes the need to plan, and it is worth paying for if that is what the buyer wants. The decision is simply being made without evidence. EV Cable Hub's 2026 home charging survey found that only 22.6% of buyers checked their own daily mileage before choosing, 51.4% chose on rated range alone and 38.1% chose on the longest single trip they could imagine making.

What the extra capacity costs is measurable in four currencies. Each additional 10 kWh costs a mean of $2,140, adds 132 lb of mass, costs 2.8% in efficiency and adds 1 hour 27 minutes to a full charge at 32A. The reasons buyers give for it are road trips at 64.2%, winter range at 42.8% and resale value at 28.1%. Against the road-trip reason sits a mean of 3.4 road trips a year, which is the same order as the 3.1 days a year above full range that the driving panel measured directly.

Asked whether they would repeat the decision, 41.4% of buyers said they would go smaller next time, 18.6% would go larger and 40.0% would buy the same again. That is an unusual distribution for a considered purchase (regret usually runs the other way) and it is the strongest single piece of evidence that the sizing decision is being made on a guess. This is the first dataset that lets it be made on a driver's own measured pattern instead, which is what the first calculator further down this page does.

Public and DC charging reliance#

US EV drivers took 19.1% of their energy from public charging in 2026 and averaged 1.9 DC fast charging sessions a month. Drivers with no home charging took 100% of their energy in public and averaged 8.6 DC sessions a month.

EV Cable Hub's 2026 panel found public reliance is set by the home setup and by almost nothing else. Drivers with Level 2 at home take 9.4% of their energy in public across 1.2 DC sessions and 0.8 public Level 2 sessions a month, at a mean monthly public spend of $18. Drivers on Level 1 at home take 21.6% across 2.4 DC sessions, at $34 a month. Drivers with no home charging take everything in public across 8.6 DC sessions and 6.1 public Level 2 sessions, at $128 a month. That is a 7.1-times spread in monthly public spend across three groups whose driving patterns are broadly similar.

The concentration figure is the one infrastructure writers should take from this page. The 16.0% of drivers with no home charging account for 44.6% of all public charging energy and 51.2% of all DC fast charging sessions. Public charging demand is therefore far more concentrated than driver counts suggest, and it is concentrated in exactly the group with the least ability to choose when they charge. Building public capacity for the average driver builds it for the wrong driver.

The cost consequence of that concentration is the clearest inequity in the dataset. EV Cable Hub's 2026 home charging survey put the mean home rate at $0.16 per kWh, the mean EV time-of-use rate at $0.11, the mean public Level 2 price at $0.34 and the mean DC fast price at $0.48. A household that can charge at home on a time-of-use plan pays less than a quarter of what a household with no driveway pays for the same energy, and the second household is on average the less able to absorb it.

Two figures on driver awareness sit underneath all of this. 44.2% of drivers are on a time-of-use or EV-specific electricity rate, and 31.6% do not know what they pay per kWh at all. Nearly a third of the market therefore cannot evaluate any charging cost claim made to them, which is worth remembering when reading any survey of what drivers believe charging costs. For drivers who do rely on the public network, our US public charging statistics study covers the network itself, and the CCS cable range covers the equipment side.

The equity finding here deserves stating without hedging. The households paying the most per mile for electricity are on average the households least able to install the equipment that would reduce it, because both facts follow from the same housing situation. A driver in an apartment pays $128 a month against $18 for a driver with Level 2 at home, on similar mileage. Every public charging price rise lands on that group and on almost nobody else, and no home tariff reform reaches them at all.

What charging actually costs by state#

Charging an EV at home cost a mean of 4.7 cents per mile across the US in 2026, from 3.2 cents in the Pacific North West to 11.4 cents in Hawaii and Alaska. Public DC fast charging cost 14.1 cents per mile nationally, three times the home figure.

The cost per mile figures follow directly from the electricity rate and the panel's measured 3.42 miles per kWh, so any reader can reproduce them. At the national mean home rate of $0.16 per kWh that is 4.7 cents a mile. On an EV time-of-use rate at $0.11 it is 3.2 cents. At the mean DC fast price of $0.48 it is 14.1 cents. Annualised across the national mean of 13,578 miles, home charging costs $638 a year, and the same driving on a time-of-use plan costs $434.

The regional spread is wider than the national average suggests and it is driven almost entirely by electricity prices rather than by driving. Hawaii and Alaska sit at $0.39 per kWh and 11.4 cents a mile for an annual home cost of $1,073. The Pacific South West runs $0.28 and 8.1 cents for $1,038 a year, and New England $0.26 and 7.5 cents for $871. At the other end the Pacific North West runs $0.11 and 3.2 cents for $412 a year, and the West North Central states $0.12 and 3.5 cents. A driver moving from the Pacific North West to Hawaii would see their charging bill multiply by more than two and a half times on identical driving.

Time-of-use is the single largest lever a driver has and it is available in every region in the table. The EV-specific rate is between 27% and 39% below the standard residential rate everywhere, and the absolute saving is largest exactly where the rates are highest: 6.9 cents a mile against 11.4 in Hawaii and Alaska, 4.9 against 8.1 in the Pacific South West. EV Cable Hub's 2026 home charging survey found only 44.2% of drivers on such a plan, which means the majority of the market is leaving the largest available saving untouched.

The comparison that matters for a household budget is between the columns rather than within them. Public DC fast charging costs more per mile in every region than home charging costs in the most expensive region in the country. A driver who cannot charge at home pays 13.4 to 18.4 cents a mile wherever they live, and the geography that dominates home costs barely matters to them at all.

Range anxiety against range use#

54.2% of US EV drivers reported experiencing range anxiety in 2026, while the same drivers used a mean of 15.8% of their available range on a typical day. Only 8.4% had ever actually run out of charge.

Setting EV Cable Hub's 2026 belief data against its measured driving data is the most direct thing this dataset does. Of the 54.2% reporting anxiety, 38.4% of all drivers report it only on long trips and 12.6% report it weekly or more often. 45.8% report none at all. Against that, the mean driver uses 15.8% of their range on a typical day, exceeds half of it 21.4 times a year and exceeds all of it 3.1 times. The anxiety is real as an experience and it is not proportionate to the risk as measured.

The behavioural figures show the same thing in what drivers actually do rather than what they say. The mean state of charge at which drivers begin looking for a charger is 26.4%, and the mean at which they actually plug in on a road trip is 18.1%. The lowest state of charge reached in the whole year averages 11.2%. 41.6% of drivers have never gone below 20% and 76.4% have never gone below 10%. The overwhelming majority of American EV drivers have never come close to the situation they describe worrying about.

The reassuring finding is that anxiety decays with experience, and it decays fast. It runs at 71.8% in year one of ownership, 52.4% in year two, 41.6% in year three and 34.2% in year four and beyond, less than half the first-year figure. EV Cable Hub's 2026 home charging survey found 84.1% of drivers would buy an electric vehicle again, and 40.0% would buy the same range again, which is the same 40.0% who told the sizing question they would repeat their battery choice.

The concerning finding sits beside it and it is not about experience at all. Anxiety runs at 46.1% among drivers with Level 2 at home, 61.4% among those on Level 1 only and 78.6% among those with no home charging. The determinant is not how far people drive or how much range they have. It is whether the car is full when they wake up. That is an infrastructure and housing finding dressed as a psychological one, and it is why the equipment sections of this page matter more to the anxiety number than the battery sections do.

Anxiety is also worth separating from its consequences, because the two are usually reported together. A driver who worries about range and consequently plugs in at 26.4% state of charge has converted the worry into a habit that costs them nothing. A driver who buys 30 kWh of extra battery for it has converted it into roughly $6,400 at the measured price per 10 kWh, plus 396 lb of mass and an 8.4% efficiency penalty carried for the life of the car. The behaviour is cheap and the purchase is not, and the survey suggests most drivers do both.

Month by month across the year#

US EV drivers covered the most ground in July at a median of 36.4 miles a day in 2026 and the least in January at 27.1. Available range moved in the opposite direction, peaking at 261 miles in September and bottoming at 224 in January.

The two series run against each other for six months of the year and that is the whole point of putting them in one table. Driving distance rises from 27.1 miles in January to 36.4 in July, a 34% increase. Available range falls from 261 miles in September to 224 in January, a 14% decrease. Utilisation therefore peaks twice: once in high summer at 17.8% in July on distance alone, and once in December at 17.0% on range alone, with December producing the most days over half range of any month at 2.2.

December is the month worth studying in EV Cable Hub's 2026 series, because it is the only one where both effects arrive together. Drivers cover 30.8 miles a day, above the November figure, while available range has already fallen to 226 miles. That combination is what produces the highest days-over-50% count of the year, and it maps directly onto the seasonal clustering of long-distance travel, where Christmas and the New Year alone hold 24.6% of all days above full range.

The last column adds the fourth effect and it is the one that gets left out. Mean home charging rate achieved falls to 6.41 kW in January and 6.44 kW in December, against 6.88 kW in September. A cold battery accepts charge more slowly, so the month with the least range and the highest heating load is also the month in which putting the energy back takes longest. Winter pressure on an electric car is a stack of four compounding effects: less range, more consumption, more distance at the holidays and slower replenishment. The category usually reports only the single cold-weather range figure.

EV Cable Hub's 2026 driving panel found the practical consequence of that stack is small for most drivers and concentrated for a few. January utilisation of 15.1% is barely different from the annual mean of 15.8%. The difference shows up in the tails, in the days over half range and in the charging time required, which is why a driver with a marginal setup notices winter and a driver with a comfortable one does not.

Seasonality also explains a reporting artefact worth watching for. A study running from spring to autumn will record daily distances between 30.6 and 36.4 miles and available range between 241 and 261, and will therefore report utilisation of roughly 15% to 17%. A study running across a northern winter will record the same drivers at 27.1 miles against 224. Neither is wrong, and the annual figures on this page exist precisely so that seasonal snapshots can be placed against a full-year baseline.

Interactive tools#

Three tools sit on this page alongside a searchable index of every figure in it and a twenty-six item readiness checklist. All four run entirely in the browser and every coefficient behind them is published in the tables above.

The two calculators are built to reproduce the published tables exactly on their default inputs, so their output can be checked against Table 19, Table 21, Table 22 and Table 25 rather than taken on trust. The state comparison tool draws on Table 2, Table 3 and Table 5, and covers all 51 jurisdictions.