The government will require new cars and trucks to meet a fleetwide average fuel economy standard of 54.5 miles per gallon by 2025.
Many motorists know they can't expect to get the mileage they see on the window sticker if they drive too fast or don't keep the car in optimal condition. But what most of them don't know is even if they do drive like highway saints, they still won't get the Environmental Protection Agency-rated results.
Under rules announced last week, tough new mileage standards will be phased in starting in 2017. But critics say those mileage ratings are notoriously unreliable, even under optimal conditions.
That's because the mileage tests are conducted with professional drivers inside of laboratories using better-performing fuel and with air conditioning turned off for most of the ride.
Harold T. Holmes thought the Chevrolet Equinox he bought recently would get the promised 32 mpg, yet when the retiree from Columbus, Ohio, began driving the small SUV, he got 26 to 27 mpg at best. And that's even with conservative driving habits such as anticipating red lights and avoiding sudden stops.
"It is all funny math and it is very frustrating when you invest $30,000 in a car and it doesn't get anywhere near what they are advertising," said Holmes, who can "squeeze mileage from a rock," in the words of his wife, Judy Wharton.
When it comes to fuel economy, drivers such as Holmes are finding the numbers don't add up. And that's why some say the debate over new fuel-economy requirements seems hollow.
Critics say the problems with the mileage tests conducted by the EPA start with testers who never break a speed limit and operate cars with far more precision than typical drivers. Vehicles are tested inside laboratories on machines called dynamometers rather than on outdoor test tracks.
The tests also don't account for jack-rabbit starts and quick stops. The highway speed portion of the test averages only 48.3 miles per hour and tops out at 60. Few drivers go less than 60 mph in open highway traffic. The test lasts about 95 minutes, but the car's air conditioning is on for just 10 minutes.
Cars the EPA tests run on indolene, a form of gasoline used specifically for the tests to avoid variations. But people filling vehicles at the corner gas station often buy fuel that is mixed with 10 percent ethanol, which has one-third less energy than gasoline.
The test procedure was mandated by Congress after the Arab oil embargo in 1973 as part of an effort to increase fuel efficiency.
Some additional tests were added later solely for the EPA to come up with more accurate window sticker ratings.
But even the window stickers don't reflect real-life driving conditions, according to the Sierra Club, which has called for overhauling the way the EPA tests mileage.
"When it comes to how we set fuel efficiency standards, we are in a world of fantasy numbers that are divorced from reality," the environmental group said.
Certainly drivers are finding that's the case. "Excessive fuel consumption" was among the most frequent complaints in this year's J.D. Power and Associates initial quality survey of new vehicles.
Ed Buchanan of San Antonio said he was "oversold" on the mileage for the Nissan Versa subcompact sedan he recently purchased to replace a gas-thirsty van.
"I was expecting to save a lot of money on gas, and it's just not there," Buchanan said.
Although they use mileage ratings prominently in advertising, automakers place any blame on the EPA.
"The government sets the test for those window stickers. We have nothing to do with what the outcome is," said Michael Albano, a Chevrolet spokesman.
For the most part, the EPA relies on automakers to conduct their own tests, based on the standards, and report the results. The agency physically tests about 15 percent of the models on the market.
The EPA said in a statement that "fuel economy for each individual vehicle will vary since no test can ever account for all driving styles, maintenance practices, and road and weather conditions."
"Unfortunately, there is no perfect test -- drivers will get higher or lower mileage," the agency said. On its website, the agency warns drivers that the tests are useful for comparing models but that the rating "may not accurately predict the average MPG you will get."