EV Winter Range Mega-Test
It was a tough time to visit Canada as an American for Greg Bannon. Greg is the AAA’s director of automotive research, who was in Ottawa, the Canadian capital, to participate in partner association CAA (Canadian Automobile Association)’s ambitious long-distance winter test drive of 14 different battery-electric vehicles in a chilly and snowy Canadian weather.
The aim of the event was to drive 14 EVs on a long and cold road trip, to carefully document how far each could travel on a full charge, in the same conditions – right up until the battery died on the side of the road.
Temperatures over the three-day event in the first week of February ranged from minus seven to minus fifteen degrees Celsius (20 to 5 °F), with a total route distance of 427 km (265 miles) of mixed city, highway and secondary (30-50 mph or so) roads, in that order.
The direct route from Ottawa to our destination hotel in the ski resort town of Mont-Tremblant, Quebec normally takes two hours via a 100-mile route, traveling northeast. But our five-hour-plus drive was carefully planned, so that even if someone finished the entire route on a single charge (as at least one EV did) there was a further loop planned to ensure on-road max range was achieved.
All routes were designed to keep electric cars that ran out of juice well past highway sections and in relatively safe areas. But knowing so many EVs were about to go down in one afternoon, there was still an element of trepidation at being stuck helpless at the side of the road, having driven to – or past – the zero-miles-left range estimator. Or “guessometer,” as early EV drivers dubbed it.
Every driver experienced some type of reduced power mode, many turtle modes were seen and driven past, until each EV could no longer move forward on its own power.
Range and Charging in the Cold
Once towed onto a flatbed by CAA’s rescue services or charged by a specially-equipped truck with DC and AC charging capabilities from mobile EV charging company CAFU, the observed distance travelled would be compared with official government EV range estimates listed on automaker sites and window stickers.
Not surprisingly, none of these EVs reached their official range figures in these conditions, with differences swinging from a fairly reasonable fourteen percent loss (Chevrolet Silverado EV and Polestar 2) to a dismal 39 percent loss for the Volvo XC40 Recharge.
“The vast difference in results highlight the importance of truth in advertising when it comes to EV range, and of comparing numbers if winter performance matters to you,” said Ian Jack, vice president of public affairs for the CAA. The national group is pushing for a new standardized labeling system for EVs, one that includes more realistic winter driving figures than just a single overall average range.
For drivers at the event, there were rules to be followed, in order to limit the number of variables as much as we could, and help make the results as comparable as possible. We were instructed to drive normally for winter conditions, with no hypermiling techniques or aggressive inputs, and asked to keep close to the speed limit, and not exceed it by more than five kilometers an hour (three mph). Which is tough in any vehicle, but especially in super quiet EVs.
Perhaps the toughest rule to follow in these conditions was not to use the heated steering wheel or heated seats, and cruise control was off-limits as well. Luckily, all EVs were allowed constant climate control use, all set to Auto mode at an even 21 degrees Celsius (70 degrees Fahrenheit).
EV Pickups Perform Well in Winter
Our 2025 Ford F-150 Lightning had no issues generating plenty of heat for its large interior. In fact, it had us removing jackets while on the move, instead of the usual turning down of the heater. Our tester was a Lariat model with the Extended Range 150 kilowatt-hour battery pack (131 kWh usable) which was rated at an official 515 km (320 mile) range.
Unfortunately, an overnight charge on a Level 2 charger was not enough time to give it a full charge by the time we had to leave. So we left with an 89 percent charge in the morning, with an estimated range of 351 km (218 miles).
Following the morning drive, after lunch is when the real fun began. This is when folks went around asking, “what are you at?”, trying to determine who was most likely to go down first, knowing that we were all heading out into the snowy, lonely landscape with a similar fate in store.
No one wanted to leave before at least one of the flatbed drivers finished their lunch, even though there were only a couple of tow trucks with our 14 EV group. So we realized we’d be depending on other CAA tow trucks in the area to rescue us from the side of the road.
This isn’t something I’ve experienced in thirteen years of owning EVs, not for battery or mechanical reasons – and even longer test-driving and covering them professionally.
The Lightning's prominent yellow warning light came on with roughly 40 km (25 miles) of predicted range left, but it still drove normally; and after clearing the light, the warnings were much more subdued than many other systems. At roughly nine percent of charge, the Lightning started losing power going up hills, and its top speed became closer to 50 mph, though that was helped by plenty of downhills as well. No paddle or extra regeneration modes were allowed to help extend range, but it was clear that both gravity and default regeneration helped here.
Which EVs Are Worst In the Winter?
The first to go down was the Toyota bZ4X SUV, not surprisingly, which turned in awful real-world cold range results, reaching 255 km or just 158 miles on a full charge and even worse DC charging performance. The Toyota brand’s only all-electric EV took the longest to charge from 10 to 80 percent, a mind-numbing 92 minutes, and averaged the lowest average DC charging speed over that time, at 33 kW.
This was slower in both time and average charge speed than the 2024 Kia Niro EV, the next-slowest in this DC fast charge test. The Niro is a smaller and less expensive EV overall, with a max DC charge speed of 72 kW, which is technically half as fast as the 150 kW DC charging capability of the 2024 bZ4X. At least on paper.
But this is why real-world tests like this are so important. Toyota itself in its support pages says that bZ4X charge times under 50 degrees Fahrenheit “will increase significantly,” and even more so for the all-wheel drive model.
“For bZ4X AWD, DC Charging will slow down further than other models when the temperature is below 32 degrees Fahrenheit and may not be possible when the temperature drops to around -4 degrees Fahrenheit or below,” a temperature we didn’t quite hit, but came very close.
By the time our Ford Lightning’s range estimator hit single-digit miles, the search for safe pull-off points became paramount. After entering a reduced power mode and slowly crawling to a stop before the crest of a hill, we noticed it had no issues going into reverse to back down into a safer spot on a less-curved part of the road.
The final observed distance for the F-150 Lightning was 296 km (184 miles), from an 89 percent charge at the start. That’s a calculated 35 percent difference between actual and its official range, which is near the back of this group here. Notably, for the eighteen minutes we waited in the Ford for the tow truck, the Lightning still produced heat, perhaps using that last remaining 1 km (0.6 miles) of range still displayed on the dash at zero percent.
The Lightning was one of three vehicles that didn’t travel beyond a zero percent state of charge reading, with the other two being the Hyundai Ioniq 5 and the Kia EV9, both 2024 models. Keep this in mind if you own or drive any of these three EVs: there’s little to no “reserve tank” to save you if you cut it a little too close.
Best and Worst Winter EV Performers
The star performer of the event was the 2024 Chevrolet Silverado EV, finishing top in multiple rankings: overall total range, quickest average charging speed for 10-80 percent DC charge (233 kW average), and it also tied for the lead in two other key rankings: percentage of range achieved, and the related percentage of range loss.
The Silverado EV’s gargantuan 205-kWh usable battery could only reach 73 percent by the time the test started. Thus, both electric pickups were the only EVs to start the day without a full charge, which meant that the actual observed range for both of these trucks was not as high as it could have been, requiring some math via its fuel efficiency average to extrapolate how its overall winter range would compare with the official figures.
And thus, an asterisk for the Silverado EV’s range figure, which was still the furthest of the group by far, at 456 km (283 miles). But that’s not so hot when you compare it with this truck’s official 724 km (450-mile) range at temperatures closer to our interior settings.
The CAA also recorded another interesting metric: the 15-minute DC quick charge test. It measured how far each EV predicted it could go from the time it hit a 20 percent charge, to the 15-minute mark. The Tesla Model 3 won here, adding 205 km (127 miles) of predicted range after a quarter hour.
Granted, the Model 3’s top showing here was helped greatly by its impressive efficiency and propensity to over-estimate its driving range (its optimistic “guessometer” is the second-lowest in predicted range accuracy).
Speaking of that metric, honorable mention should go to Kia the company as well for its range prediction calibration. The small Niro and large EV9 three-row SUV took gold and silver at actually travelling the closest to their predicted range estimates, achieving 99 percent and 97 percent respectively of what the dashboard said that morning.
The worst in this key “Displayed Range Achieved” metric? The Chevrolet Equinox EV, surprisingly, delivered just 66 percent of its promised range.
Easily the second-best performer here overall was the Polestar 2, which tied the Chevy pickup with the least difference between official range (444 km or 276 miles) and observed range (384 km or 239 miles), with only a 14 percent loss. But the Polestar 2’s numbers also deserve an asterisk, as its driver kept the heat off in the car until the lunch stop by accident, a major efficiency advantage in the cold.
Conclusion: Real-World Winter EV Range Test
An engineer by trade, the AAA’s Bannon says his group also did a cold weather EV test in 2019, but used a much more scientific methodology based on the one used by the EPA, where vehicles were brought onto a dyno in a lab that could be cooled to winter temperatures, which could standardize wind, pedal inputs and traffic conditions. That test six years ago on first-gen EVs found that at 20 degrees F, the average driving range “decreased by 41 percent,” when using the HVAC system to heat the vehicle, as in this test as well.
This very different test’s real-world methodology is for sure less scientific, but came in at an average range loss of roughly 25 percent compared to its official government/sticker EV range for the entire group of vehicles.
The EPA came down on automakers years ago to make gasoline fuel efficiency figures more accurate. Perhaps to increase confidence in all-electric vehicles, especially in northern and colder states, maybe it would be better to spell out their true capabilities and limitations in winter for all to plainly see, instead of hiding behind ideal ranges at ideal temperatures.
