What gets the better gas mileage: windows up, AC on, or windows down, AC off?

Hey i just read the post on driving with AC on vs AC off. I’m a huge Myth Busters fan and they actually came to the resolve that at a certain speed it becomes more efficient to have windows up AC on. If i recall correctly it was at about 45 miles per hour. Do you guys agree or disagree?

What gets the better gas mileage: windows up, AC on, or windows down, AC off?, I rather fancy.

As to the point itself, as the original column indicates, there are any number of variables here – the aerodynamics of the car with open windows, the aerodynamics of the car with closed windows, the relative load of the AC on the engine, the ambient temperature, and even the color and the amount of glass in the car. The Mythbusters result seems to lie between the two extremes we already have.

Depends on car shape (make, model, and year dependent), car color, engine and other options, speed, hillyness (pretend that’s a real word), amount of sunlight, humidity, precipitation, traffic, and maybe a few others. Other than that, it’s a simple problem. Some cars will always go one way, some the other, and some in between. Your mileage may vary, consult dealer for details.

The only way to tell for sure on your car is to try it. Mine usually likes AC on.

I generally love Mythbusters, but I have to say, that was one of the worst MB experiments ever. They compared two cars that were identical in every way; they had one car close its windows and crank the A/C up to maximum. I think it was Jamie in that car. He was wearing a parka and practically had icicles growing in his mustache by the end of the test. The other car (driven, IIRC, by Adam) just cracked the windows and drove around the track. On a temperate San Fransisco afternoon when it was probably 75 degrees outside.

This is not an apples-to-apples comparison. Of course cracking the windows is more efficient than dropping the temperature of all the air in your SUV down to 45 degrees. But for normal circumstances – say it’s 85 outside and you want it to be 75 in the car – the MB experimental data is totally worthless.