You can take it to the pump: Your vehicle's mileage will vary from most miles-per-gallon ratings published by the Environmental Protection Agency.

It's not a conspiracy. The government isn't lying. In a quixotic pursuit of normalcy, it's simply conducting mileage tests in a world that doesn't exist.

First, the government uses professional drivers, which automatically skews testing outcomes. How many professional drivers are on the road at any given moment? Can you spot one? Is it the person meditating in the left lane of an interstate highway, poking along, forcing other drivers to use more fuel and take more risks by accelerating and passing on the right? Is it the person exceeding the posted speed limit by 10 to 20 miles per hour? Is it you?

Roads have a big effect on fuel economy. Rough roads give your vehicle and its tires a more vigorous workout. Smooth roads favor better mileage. Why? The more anything works, the more energy it uses. So, let's go back to those professional drivers.

To be exact, some of those drivers work for the car companies and others are employed by the government. The company drivers and their bosses run mileage tests required by the EPA. They test pre-production models of vehicles and pass the results along to the government.

The EPA's drivers run the same tests on about 10 percent of the company's models to check the company's results. If the government's results match the company's results, the vehicles get an EPA mileage sticker with the consumer caveat: "Your actual mileage may vary." No wonder.

Look at the tested vehicles. They are "pre-production" models. That's industry parlance for "not ready for market," as in, if something is wrong, it will be fixed before the regular, retail-ready production run. Those fixes can be big or small, but they can all affect actual mileage.

How are the actual tests conducted? Here is the EPA's verbatim response in its frequently asked questions document, available at www.fueleconomy.gov: "The vehicles are driven by a professional driver under controlled laboratory conditions, on an instrument similar to a treadmill."

That instrument is called a dynamometer (dyna-mom-e-ter), which is a device for measuring force or power. It does not have potholes, ruffles, ridges, rough surfaces or anything else you might find on a real road; and, as noted earlier, all of those variations affect actual mileage.

But what good is a test result if it can't be repeated? How can you get repetition if variables are introduced into every test? You can't, as the EPA acknowledged in this statement: "These procedures ensure that each vehicle is tested under identical conditions; therefore, the results can be compared with confidence."

The problem, of course, is that the real world is the home of variability. There, the only thing certain is uncertainty. There, difference, and lots of it, is normal.

The EPA seeks two different fuel-economy estimates for its tests, one for city driving and another for highway motoring. Thus, several separate tests are required.

Of note are the basic city and highway protocols. The city test simulates an 11-mile urban run, replete with stop-and-go traffic, at an average speed of 20 miles per hour. The "trip" takes 31 minutes and includes 23 stops.

Come on. I live on a quiet street in Northern Virginia where respect for the 25-mph speed limit is the exception, rather than the rule -- especially, oddly enough, when local schools are in session. And the nine-mile trip from my house to The Washington Post's office downtown can take substantially longer than 31 minutes depending on the time of day. And what about those poor souls doomed to traverse the Pacific Coast Highway in the Los Angeles area? If we could find a way to retrieve all of the fuel wasted through idling in PCH traffic jams, we could cut the crude umbilical cord that ties us to the Middle East in so many unpleasant ways.

What about the highway test? It simulates a 10-mile trip at an average speed of 48 miles per hour, with a maximum speed of 60 miles per hour. There are no stops along the way. There is no one messing around in the left lane, the passing lane; no aggressive drivers, no 18-wheelers, no SUVs, no econoboxes masquerading as sports cars, none of that stuff.

I have no desire to drive that EPA highway anytime soon. It sounds frightfully like the Highway to Heaven.

EPA mileage ratings tend to be optimistic, mostly because cars aren't tested in real-world conditions. This one is for a Honda Civic Hybrid. Hybrids often get better mileage in the city.