Consumers tend to think that the large black numbers on a car's window sticker, which list the EPA estimated fuel economy rating, are what they will actually get while driving that car. Unfortunately, this is another situation covered by the saying, "Read the fine print." The fine print says, in essence, actual mileage may vary depending on a whole lot of thingsWhat you as a car shopper need to know is that the gas mileage you get could be less. A lot less. Before you begin hatching some conspiracy theory involving oil companies, the U.S. government and carmakers, you should take a look at where these mileage estimates come from.

The government puts a lot of time and care into testing for fuel economy. And these estimates provide important information for comparing two vehicles. What critics say is that, as part of the test program, a number of assumptions were made about driving habits. Those assumptions are now outdated; hence the discrepancy.

New vehicles are tested by the manufacturer according to guidelines outlined by the EPA. The EPA in turn reviews the results and confirms 10 percent of the test results with additional testing in the National Vehicle and Fuel Emissions Laboratory (NVFEL), in Ann Arbor, Michigan. The cars are tested by putting the drive wheels on a "dynamometer," a machine that allows it to simulate driving conditions while remaining stationary. The car is tested according to two different "schedules" to produce the city and highway ratings posted on a car's window sticker. (Fuel economy ratings are also included on the "Specs" page of all the cars on Edmunds.com.) The two schedules are meant to simulate different styles of driving that would be encountered in the two types of driving:

City: The test is started with the car's engine cold. The car then is operated in such a way as to replicate 11 miles of stop-and-go rush-hour traffic. Periods of idling are included in the test; average speed is 20 mph and top speed is 56 mph.