Fuel economy has been in the headlines recently.

There are new standards coming that will lower the overall average of all new cars and trucks by 2025 to the current level of a Toyota Prius.

And there are vehicles that aren’t living up to the fuel economy their manufacturers have officially claimed.

But do you know how fuel consumption is actually calculated? It may surprise you, especially since there’s no driving involved—and they don’t even measure the fuel itself.

The testing itself is done by the automakers, using a standardized procedure approved by Transport Canada, which is responsible for ensuring that the data is accurate. The results are submitted to Natural Resources Canada, which bundles them all into its annual Fuel Consumption Guide.

You may have thought the testing was done on a track, on a perfect day, with the lightest person on the team behind the wheel. In reality, testing is done in a laboratory. You might argue that this detracts from the “real-world” aspect of the test, but it’s the only way to ensure consistency, which is essential to the process.

Each vehicle has about 6,000 kilometres on it when it’s tested, using a two-wheel dynamometer. Vehicles with four-wheel or all-wheel drive have the system disconnected and are tested in two-wheel only. The ratings are adjusted and corrected for such things as aerodynamics, rolling resistance, weight, and if applicable, the four-wheel system.

Other factors such as options and configurations are also included in the final calculation. (What tripped up Hyundai and Kia in its miscalculated fuel figures was a test that incorrectly measured the rolling resistance, friction loss and aerodynamic drag, required to compute the final figure.)

The city driving segment takes 23 minutes and covers a simulated 12 kilometres. The vehicle makes 18 complete stops, with four minutes spent idling to represent a driver waiting at traffic lights. The top speed is 91 km/h, while the entire test averages 32 km/h. The test is initially done with the engine temperature similar to starting the car after sitting overnight in the summer.

A portion is repeated with the engine restarted after reaching optimum temperature and then sitting for a short time.

The highway portion takes 13 minutes and covers 16 kilometres, but with no stops. The highest speed reached is 97 km/h, with the overall average speed set at 77 km/h, to represent highway and rural roads.

As mentioned, the testers don’t measure the amount of fuel in the tank. Rather, they measure the amount of carbon in the exhaust, which they use to determine how much fuel was burned. This method works for any carbon-based fuel, which includes gasoline, diesel, ethanol, and natural gas.

Occasionally, the manufacturers’ results will be audited by Transport Canada. This is done if the numbers seem out of whack, and may also be done on models that haven’t been tested in a long time, or have new technologies. These tests are done on models that the government buys from dealerships.

The vehicle is driven 3,500 km at specified speeds, and then filled with a standardized test fuel for a laboratory test. As with the automakers’ testing, the emissions are analyzed to get the final figures.

Testing in the U.S. is also done in a laboratory environment, but includes some extra “real-world” tests, which started with the 2008 model year. This is why American fuel economy ratings differ from Canadian ones even after you’ve accounted for the different sizes of the U.S. and Canadian gallon.

The city and highway tests are pretty close to ours: there is a cold start for the city, the test is run, and then a portion is retested with a hot engine, averaging 21.2 mph (34 km/h). The highway test uses a warmed-up engine at an average 48.3 mph (78 km/h).

But the Environmental Protection Agency (EPA), which oversees the process, adds other tests. There’s a “high speed” test, which still averages 48.3 mph, but with more aggressive acceleration and braking; an air conditioning test, in which the car is driven on the dynamometer with the a/c on and the ambient temperature at 95F (35C); and one in which the vehicle is driven in simulated stop-and-go traffic, from a cold start, in 20F (-7C) ambient temperature.

Will we ever see this more realistic, five-cycle test done in Canada? According to Transport Canada’s website, “The Government of Canada will analyze this test procedure to determine its feasibility in a Canadian environment.” Given that the U.S. announced its upcoming new methods in January 2006, we’ll suggest the translation is, “Don’t hold your breath.”

Currently, fuel economy ratings are only for light-duty vehicles. Trucks and SUVs with a gross vehicle weight rating (GVWR) of more than 4,536 kilograms (10,000 lbs) are exempt from testing. Also exempt are any cars or crossovers with a GVWR of more than 3,856 kg (8,500 lbs) or a curb weight of more than 2,722 kg (6,000 lb).

You’ll now also notice a new rating, Le/100 km (in the U.S., it’s MPGe), which stands for litres equivalent. It’s used for electric cars, and for range-extended and plug-in hybrid vehicles when they’re running solely on electricity. It’s an equivalency measurement of what the electric vehicle would get if it were running on gasoline.

It’s not ideal, but it gives buyers some idea of how these cars compare to conventional vehicles when they’re making a decision.

So why can’t you get the published fuel figures in your own vehicle? They can be achievable, but there are numerous factors. Some of these will include cold ambient temperature and how much you use the air conditioning, but the most important one is the driver.

Hard acceleration and braking, speeding, and jerky throttle pressure, along with not maintaining your vehicle, will always cost you more at the pump.