US Electricity Is About 18.3¢ per kWh in 2026 — What That Does to an Appliance Bill
2026-09-18
Short answer: The US residential average is about 18.3¢ per kWh (EIA, September 2026), up roughly 5% on a year earlier. Running cost = watts ÷ 1,000 × hours × rate — so a 1,500 W heater run six hours a day for a 120-day winter costs about $198 at that rate, and more in a state above the average.
The figure
The US residential average retail price of electricity is about 18.3 cents per kilowatt-hour as of September 2026, roughly 5% higher than a year earlier. The Energy Information Administration publishes this monthly in the Electric Power Monthly series, by state and by sector.
Two things about that number matter more than the number itself.
It is an average of averages. Residential rates range from under 12 cents in some states to over 30 cents in others. Using 18.3¢ to estimate your own bill in a high-rate state understates it badly, and in a low-rate state overstates it.
It is an energy rate, not a bill. Most residential bills add a fixed monthly customer charge and, in many markets, separate delivery or distribution charges. The rate that matters for “what does one more hour of this appliance cost” is the per-kWh energy rate; the rate that matters for “what will my bill be” is the total divided by the kilowatt-hours.
The arithmetic
running cost = (rated watts ÷ 1,000) × hours used × price per kWh
Worked example — a resistive space heater, the appliance this arithmetic is most reliable for, because it draws its rated power the whole time it is on.
- Rated power: 1,500 W → 1.5 kW
- Use: 6 hours a day, 120 days of the heating season = 720 hours
- Energy: 1.5 × 720 = 1,080 kWh
- Cost at 18.3¢: 1,080 ×
0.183 = **197.64**
At 12¢ that same heater costs 129.60 for the season. At 30¢ it costs 324. The heater did not change; the state did. This is why a cost-per-hour figure copied from an article is worth less than the same arithmetic run on the rate printed on your own bill.
Where the nameplate lies
The formula above is exact for anything that runs at full power whenever it is switched on: kettles, toasters, hair dryers, resistive heaters, incandescent lighting.
It is wrong — often by a factor of three or more — for anything thermostatically controlled. A fridge rated 150 W does not use 150 W for 24 hours; its compressor cycles, and its real consumption is closer to its energy-label annual figure of a few hundred kilowatt-hours a year. The same applies to air conditioners, heat pumps, freezers and anything with a duty cycle.
For those appliances, use one of these instead of the rating plate:
- The annual kWh figure on the energy label, divided by the hours you actually run it.
- A plug-in energy meter, left in place for a full week of normal use.
- Your own bill, compared before and after a month with the appliance unplugged.
What would reverse the conclusion
A 5% rate rise is small in percentage terms and can still be decisive at the margins:
- On an efficiency upgrade, a higher rate shortens the payback period proportionally. A 5% rise turns a 10-year payback into about 9.5 years. If a decision hinged on that difference, it was too close to call either way.
- On an always-on load, the rise is pure recurring cost. At 1,000 kWh a month, 5% is about
11 a month — small monthly,132 a year, and permanent. - On switching fuels, it moves the line directly. Electric resistance heating competes with gas on a price ratio, so an electricity rise and a flat gas price both push the same way; a heat pump, which delivers several units of heat per unit of electricity, is far less sensitive to the same rise.
Run your own numbers in the appliance running cost calculator →