Your mileage may vary. Every driver has read this disclaimer and has learned to take automotive showroom window-sticker fuel-economy estimates with a grain of salt.

The window stickers are required by law to give new-car buyers a sense of how much they’ll have to spend on gas. A decade ago, Consumer Reports tests indicated that the EPA fuel economy estimates on window stickers were off by 10 percent or more. Since then, the Environmental Protection Agency (EPA) has revised its procedures, adding in the use of air conditioning and driving cars at higher speeds to better reflect how U.S. drivers actually drive. 

A new Consumer Reports analysis shows that the current EPA window stickers are pretty close to what you can expect to get in the real world. Examining 399 models from 2009 through 2016, we found a 3-percent variation between EPA values and what we found in our own testing. The mileage reported by respondents to Consumer Reports surveys has long tracked with our testing, and now it reflects EPA window stickers, too.

"Consumers should be able to trust that the estimate they see on the label accurately reflects their gas mileage,’’ says Shannon Baker-Branstetter, policy counsel for Consumers Union, the policy and mobilization arm of Consumer Reports. "We’re glad that today, consumers are getting more accurate information that they did in the past.’’

EPA Fuel Economy Shown on Window Sticker

By law, the EPA conducts its fuel economy tests under lab conditions to ensure it gets the same results every time. Cars are tested inside on a dynamometer, according to a carefully proscribed set of simulated driving maneuvers. In 2008, to more closely model actual driving behavior, the agency added some faster speeds, turned on air conditioning and made the external air temperature colder. Automakers tests cars in their own test facilities using EPA procedures. The agency does spot checks on 15 to 20 percent of models offered for sale at the National Vehicle and Fuel Emissions Laboratory in Ann Arbor, Michigan.

Consumer Reports tests are also designed to minimize variability. Our vehicles are parked overnight in a garage before testing. Tire pressures are set to manufacturer's’ specifications. The fuel tank is full at the start of each trial. All testing is done when temperatures are above 32 degrees (50 degrees for hybrid and electric vehicles), with winds at 15 mph or less and no precipitation. A fuel-flow meter with a readout in the cabin is used to measure consumption. Results are adjusted for temperature variations.

Our city tests are performed along a 1-mile course precisely marked out at our Connecticut test track. The stop-and-go city-driving simulation has three stops, includes 40 seconds of idle time and speeds never exceed 40 mph. Our highway tests are done on a specific section of Connecticut Route 2 near the test track. The course is exactly five miles long, and cars are driven at a constant speed of 65 mph. (Learn more about how Consumer Reports tests cars.)

The biggest discrepancies between Consumer Reports tests and EPA tests now are in gas-electric hybrids. There was a 9 percent difference between the window-sticker value and what we observed. Our survey respondents reported mileage closer to the EPA’s lab tests—about 3.6 percent greater than the window sticker numbers. That could be because hybrid owners tend to drive their cars in ways that maximize fuel economy—something that wouldn’t be reflected in standardized testing, says Jake Fisher, director of automotive testing at Consumer Reports.

Consumer Reports will continue its fuel-economy tests to ensure automakers are living up to their promises and flagging models that don’t seem to deliver, as we did when Ford was overstating mileage claims on its 2013 hybrids.

"While Consumer Reports appreciates that EPA window sticker numbers have improved, we will continue to test to identify anomalies that are not revealed in a trend analysis,’’ Fisher says.

Having reliable numbers is important to drive the market for efficient cars, says Christopher Grundler, director of EPA’s Office of Transportation and Air Quality, which administers the window-sticker tests. The key to EPA’s fuel-economy labels is that all vehicle models are measured in the very same way, so different models can be compared, he said.

“We’re happy but not surprised to hear that EPA’s fuel economy labels are a good reflection of what people are getting in the real world,” Grundler says. “Higher fuel economy saves drivers money at the pump and protects the climate by emitting less of the carbon pollution that contributes to climate change.”