Five ways of saying the same thing, and one that is different
Every efficiency figure on this page except MPGe is the same measurement wearing a different hat. kWh per mile is the raw quantity: energy divided by distance. Wh per mile is that number times a thousand, which puts it in a range humans read easily — 250 to 400 for most passenger EVs. kWh per 100 miles is Wh/mi divided by ten, and it is what appears on the US window sticker because it behaves like the litres-per-100-km convention the rest of the world uses. Miles per kWh is the reciprocal, and it is the one that misleads.
Reciprocals are dangerous because equal steps are not equal savings. Going from 2 to 3 mi/kWh saves 500 − 333 = 167 Wh/mi. Going from 4 to 5 mi/kWh saves only 250 − 200 = 50 Wh/mi, a third as much, despite being the same one-unit step. That is the identical trap that makes mpg misleading in gasoline vehicles, and it is why regulators prefer the consumption form.
MPGe is genuinely different, because it is defined on energy taken from the wall rather than from the battery. The EPA assigns a gallon of gasoline an energy content of 33.7 kWh and then asks how far the vehicle goes on that much metered electricity. Charging losses therefore count against MPGe but do not appear in the car's own trip computer, which is why a car showing 3.5 mi/kWh on its display does not post 118 MPGe on its label.
Efficiency is what actually determines cost and range. Battery capacity buys range only in proportion to it: range is usable kWh multiplied by miles per kWh. A 75 kWh pack at 3.4 mi/kWh and a 90 kWh pack at 2.8 mi/kWh deliver 255 and 252 miles respectively. Pair this with the EV charging cost calculator to turn Wh/mi into cents per mile.
Getting the two energy figures right
The whole calculation hinges on knowing where the energy was measured, and the two answers differ by the charging efficiency.
Battery energy is what the car's trip computer reports. It is the energy that left the pack and went into the motors, accessories and cabin. It is the right basis for comparing driving efficiency between trips, because it excludes everything that happened while the car was parked.
Wall energy is what the charger or the utility meter records. It includes the onboard charger's conversion losses, resistive losses in the cable and connector, and any energy the car spent conditioning the battery during the charge. It is the figure you are billed for, and it is the basis for MPGe.
Charging efficiency is battery energy divided by wall energy. It is not a single number: AC charging at low current is less efficient than at high current because fixed overheads are spread over less energy, and cold-weather charging can be markedly worse because the pack heater runs off the same supply. Measuring it yourself is straightforward — record the wall meter over a full charge and compare it with the change in the car's reported battery energy.
Distance must cover the same interval as the energy reading. Reset the trip meter and the energy display together, and use a full charge-to-charge cycle rather than a single leg where possible, since partial-trip energy displays often exclude preconditioning and standby draw.
Usable battery capacity is not the number in the brochure. Manufacturers reserve a buffer at both ends of the state-of-charge window to protect cell life, so the energy actually deliverable between a full charge and the car stopping is smaller than the gross pack capacity, sometimes by several kWh. Use a measured full-to-empty figure if you have one.
Worked example: 210 miles on 62 kWh
A trip computer shows 62 kWh used over 210 miles. The car has a 75 kWh usable pack and charges at 88% efficiency. You want every efficiency measure and a realistic winter range.
- Energy per mile. 62 ÷ 210 = 0.29524 kWh/mi.
- Watt-hours per mile. 0.29524 × 1,000 = 295.2 Wh/mi.
- Miles per kWh. 1 ÷ 0.29524 = 3.387 mi/kWh.
- kWh per 100 miles. 0.29524 × 100 = 29.52 kWh, the figure that would appear on a window sticker.
- Wall energy. 62 ÷ 0.88 = 70.45 kWh. The meter records 8.45 kWh more than the battery received.
- MPGe. 210 × 33.7 ÷ 70.45 = 7,077 ÷ 70.45 = 100.4 MPGe.
- Projected range. 75 × 3.387 = 254.0 miles.
- After a 25% derate. 254.0 × 0.75 = 190.5 miles.
Two comparisons are worth drawing out. Had you computed MPGe from the battery figure instead of the wall figure, you would get 210 × 33.7 ÷ 62 = 114.1 MPGe — 13.6% higher than the correct 100.4, which is exactly the reciprocal of the 88% charging efficiency, since 1 ÷ 0.88 = 1.1364. And the 25% derate costs 63.5 miles, which is the difference between arriving with margin and arriving on a flatbed.
Reading the number, and what moves it
Wh per mile is the most useful single figure because it is linear in energy and therefore linear in cost. At 295 Wh/mi and 15 cents per kWh at the meter, a mile costs 295 ÷ 1,000 ÷ 0.88 × 0.15 = 5.03 cents. Halving consumption halves the cost; the reciprocal measures do not behave that way.
Speed dominates on the highway. Aerodynamic drag rises with the square of speed and the power required with the cube, so highway consumption is far more speed-sensitive than city consumption. Unlike a combustion car, an EV does not have a broad efficiency peak at moderate cruise — its motor is efficient across the range, so nothing offsets the aerodynamic penalty. The drag and road load calculator quantifies the power side of that directly.
Cold weather is the other large term, and it acts through three channels at once: cabin heating draws real power with no waste heat to recycle, battery chemistry is less willing when cold so more energy is lost internally, and the car may spend energy warming the pack. A derate of 20–40% in genuinely cold conditions is a realistic planning allowance, which is what the derate input exists for.
Regeneration flatters short trips on descents. A leg that loses elevation can post an implausibly low figure, sometimes below 200 Wh/mi, and the return leg will pay it back with interest. Always compute efficiency over a round trip or a full charge cycle when you are planning from it.
MPGe is for comparing across fuels, not for planning. It exists so a window sticker can put an EV next to a gasoline car on one axis. For deciding whether you can reach the next charger, use miles per kWh and your usable capacity; for deciding what a mile costs, use Wh/mi and your electricity rate. Feed both into the total cost of ownership calculator when comparing an EV against a combustion alternative, and use the fuel economy calculator for the other side of that comparison.
Every efficiency measure side by side
| Wh/mi | kWh/100 mi | mi/kWh | MPGe (no charging loss) | Range on 75 kWh (mi) |
|---|---|---|---|---|
| 200 | 20.0 | 5.000 | 168.5 | 375 |
| 250 | 25.0 | 4.000 | 134.8 | 300 |
| 270 | 27.0 | 3.704 | 124.8 | 278 |
| 300 | 30.0 | 3.333 | 112.3 | 250 |
| 337 | 33.7 | 2.967 | 100.0 | 223 |
| 350 | 35.0 | 2.857 | 96.3 | 214 |
| 400 | 40.0 | 2.500 | 84.3 | 188 |
| 450 | 45.0 | 2.222 | 74.9 | 167 |
The 337 Wh/mi row is the definition point: 33.7 kWh per 100 miles from the wall is exactly 100 MPGe.
Charging losses are real energy you pay for
The gap between the wall meter and the car's display is not a rounding artefact. On AC charging it commonly represents a tenth or more of the energy delivered, spent in the onboard charger's conversion, in cable resistance and in running the thermal management system while the car sits plugged in. It is worse at low charging currents, because the fixed overheads are spread across less energy — a 120 V household outlet is meaningfully less efficient than a 240 V circuit for the same kWh delivered. DC fast charging bypasses the onboard charger but adds conversion losses at the station and usually more thermal management. Measure your own by comparing a full charge on your utility meter against the car's reported energy gain, and use that number here rather than a default.
Mistakes that produce a misleading efficiency figure
- Mixing wall energy with a trip-computer distance and calling it vehicle efficiency. That number is neither the car's consumption nor its MPGe; it is a hybrid of the two. Pick a basis and stay on it.
- Using gross pack capacity for range. The usable window is smaller, sometimes by several kWh, because the car reserves buffers at both ends of the state of charge.
- Computing efficiency over a one-way descent. Regeneration makes the outbound leg look excellent and the return leg terrible. Use a round trip or a full charge cycle.
- Comparing MPGe figures computed on different bases. The EPA figure includes charging losses. A number derived from a car's own display does not, and runs high by the reciprocal of the charging efficiency.
- Treating a summer figure as a winter figure. Cabin heating, cold cells and pack conditioning together can cost 20–40% in genuinely cold weather, which is what the derate input is for.
- Ignoring standby and preconditioning draw. Energy used while parked — cabin preconditioning, sentry modes, battery maintenance — appears on your bill and often not in the trip computer's per-mile figure.
- Planning to zero. Range projections assume constant conditions and a charger that works when you arrive. Leave a reserve beyond the derated figure.
How the official figures are produced
US EPA range and efficiency figures for battery electric vehicles come from a laboratory procedure built on SAE J1634, run on a chassis dynamometer over standardised drive cycles from a full charge until the vehicle can no longer follow the trace. Energy is measured from the wall over the subsequent full recharge, which is precisely why the label figure includes charging losses. The result is then adjusted downward by a correction factor before it appears on the window sticker, because the raw cycle results are known to be optimistic relative to real-world driving.
The 33.7 kWh gallon-of-gasoline-equivalent is a regulatory definition, not a measurement of your car. It exists so that vehicles running on different energy carriers can be compared on one axis, and it makes no claim that a gallon of gasoline would move a car as far as 33.7 kWh would — internal combustion converts only a fraction of that energy into motion, which is exactly why EV MPGe figures are numerically so much higher than gasoline mpg.
Other regions use different conventions. The WLTP procedure used across Europe and much of Asia reports consumption in Wh/km, generally measured from the wall as well, and its cycles are gentler than the EPA's adjusted figures, so a WLTP range for the same vehicle is typically higher than the EPA number. Comparing a WLTP range against an EPA range is comparing two different tests, not two different cars.
For your own purposes, none of that matters as much as your own measurement. Record energy and distance over several full cycles in the conditions you actually drive in, take the average, and use that. It is the only figure that describes your car, your route and your right foot.
Key terms
- MPGe
- Miles per gallon equivalent. Distance travelled per 33.7 kWh of energy taken from the wall, the EPA's cross-fuel comparison metric.
- Gallon of gasoline equivalent (GGE)
- 33.7 kWh, the energy content the EPA assigns to one gallon of gasoline for labelling purposes.
- Usable capacity
- The energy a battery pack will actually deliver between a full charge and the vehicle stopping, excluding the reserved buffers at each end of the state-of-charge window.
- Charging efficiency
- Energy that reaches the battery divided by energy drawn from the wall. Lower at low charging currents and in cold weather.
- Wh/mi
- Watt-hours per mile — the consumption form of efficiency, linear in energy and therefore linear in cost.
- Derate
- A planning allowance applied to a range projection to account for conditions harsher than those the efficiency was measured in.