Air conditioning and gasoline mileage


Does the use of air conditioning have any significant affect on gas mileage? When I complained about lower than expected gas mileage, the dealer tried to tell me that air conditioning uses a considerable amount of gas.

I know it uses at least some gas, but probably not enough to make a difference. I suspect that air conditioning going full blast and full time on the hottest of days probably wouldn’t burn more than a quart of an 18-gallon tank.




The average AC system uses about five horsepower to operate. If this five horsepower is no longer being used to power the vehicle down the road, it will require more fuel to makeup for this loss of horsepower if you want to keep the vehicle moving at the same rate as if the AC wasn’t being used. And this requires more fuel to do this.



The dealer is right, to some extent, that using A/C consumes more fuel than driving with your windows up and no A/C. But no one has, to my knowledge, calculated how much fuel is used or the effect on a car’s mpg. They’re still arguing about whether it’s more costly to use A/C on a hot day or to turn off the A/C and roll down all the windows. The latter method noticeably increases air resistance.

Anyhoo, your below-expected gas mileage is probably not related to A/C use. Maybe the EPA even factors that into their estimates, who knows. The dealer had to make up some baloney to feed you. We might have some further suggestions if you had supplied data (they never do) but you asked only about use of A/C. So I’ll stop right here.


I agree. I should have furnished more information.

It’s 2007 Honda Accord, 4 cyl, automatic transmission. EPA estimate is between 20 and 28 mpg in the city and 28 and 40 mpg on the highway. Ha! I’ve been carefully recording the last five fill-ups and totally topping off for accuracy. It’s getting an average of 16.7 mpg, all in town. I haven’t had occasion to use up a tank on the Interstate.

The station I do business with is a “Pure” station, a little known brand. I filled up last time at a BP station. I don’t expect it to make any difference because it’s said that all gasoline comes out of the same pot.

Thanks to all for comments.

It varies greatly from one car to another.  In most modern cars the A/C has little but measurable effect on the gas mileage. Likewise opening a window at highway speeds also has a negative effect on mileage.  

Assuming you are not talking about a 2008 car, you should not be disappointed with mileage less than the Sticker showed.  Those stickers have been very optimistic for years.  The 2008 sticker mileage is based on different methods and should be closer to what you should expect.  You might want to check and see what the 2008 estimated mileage is for your car.

Also personal driving style can greatly effect your personal mileage.  I generally have gotten better mileage that the sticker value.  In fact I have done better on every car I have had that had an estimated mileage since 1970.  Of course most people would say I tend to drive like a little old lady. 

Good Luck


This is a new vehicle. And new vehicle engines take a while before they get broke in to where all the moving parts get happy with each other. Until this happens, the mileage realized will be lower than what the EPA sticker says. So give it some time for the engine to break in.



OK with more information I would say my original comments are valid.

According to the EPA

the old numbers were:
24 city 34 highway 28 combined

The new numbers are:
21 city 31 highway 28 combined

Based on the numbers for 16 vehicles the average was 24.8 with a high of 30 and a low of 15.

That would put you on the low end of the group. If you are driving cconservatively then you are low but if you are driving aggressively you are certainly within normal range.


If the gasoline you’ve been buying has 10% ethanol blended in, it could account for the low mileage. I think 16.7 is pretty low for a 4 cyl Accord, but if your “in town” driving is mostly sitting at red lights, that could be it, too.

My 4-cylinder Accord returns about 27 mpg in mixed city-suburban driving, and more than 30 mpg on the highway.


OK, I have a 2006 Accord with a V6. I run the AC all the time and have never gotten less than 20 MPG around town. Do you have a lot of stop-and-go traffic where you live? I get around 30 MPG on highway trips. I can’t see how a 4 cyl. would do that much worse unless you have a lot of stop-and-go, or you really like to accelerate fast.

I do hear a lot of people complaining about not getting anywhere near the EPA mileage estimates. For myself, I have almost always gotten within one or two MPG of the EPA sticker in several different cars. So, apparently driving style and location must have a large effect.


16 MPG in town could very well be driving habits. It’s hard for most people to be subjective about their own habits.
The only true indicator IMHO would be on extended road trip at the posted speed limit.


Mythbusters did a segment on this once. The car running ac did a little better than the one without, if I recall correctly.
Keep in mind, 2 different drivers in similar cars on a track with a measured amount of fuel.
There are lots of real world variables that affect mileage that can’t be tested on a track.
I was rather pleased with 40 mpg @ 70 mph (ac on)in my civic until I used gps & found out 70 on the speedo was 65 in the real world. The same difference applies to the odometer, but roughly 38.5 isn’t bad either.
My point is, how nitpicky do you want to get determining the difference?

I think your mileage will continue to improve over the next 6 months to a year as the real mechanics here have implied.

Btw, are you the only driver of this car? I bought my civic used & was looking forward to at least 34 around town on the first tank. It got less, and I later found out my girlfriend was used to getting about 20 mph out of each gear, at least in 1st & 2nd…


Running the A/C has to have an effect on mileage. The compressor is drawing power from the engine, and the engine has to burn gas to provide that power. That said, in modern cars it’s usually noticeable, but not too bad. Your driving environment (lots of stop and go versus highway) and driving habits (if you’re a butthead who mashes the accelerator pedal) will probably have greater effects on mileage.

A while back, Clank and Clunk stated that careful testing had suggested the following cooling strategy: around town, below highway speeds (say, under 40 or 45), turn the A/C off and open the windows (and use the fan). The wind drag penalty of having windows open will be more than offset by the savings of not running the A/C compressor. At highway speeds (40 or 45 and up), close the windows and use the A/C if necessary. The A/C compressor power draw is less than the wind drag penalty at higher speeds.


Everybody talks about the EPA mileage estimates; but, WHAT are they based on? I went for a quick scan at EPA’s web site to get an idea. The site said the criteria is set by Federal law. I didn’t read THAT, yet. It stands to reason (mine, at least) that the test would be based on speeds WITHIN the posted speed limits. The speeds I see driven, by most people, are 10 mph, 20 mph, 30 mph above whatever the POSTED speed limits. These same people, I’m sure, complain about not getting the EPA estimated mileage! DUH! This web site covers mpg, comparisons among vehicles, testing, saving fuel methods, and so much more:


[b]Quote: “and totally topping off for accuracy”[b]

Well, if you continue to do that, then you will soon have a lot more to worry about than a drop in gas mileage. Those who insist on continuing to pump gas into the tank after the first “click-off” are OVERFILLING the tank to the extent that the excess can pollute the carbon canister that is the heart of the evaporative emissions system.

If you stop this foolish practice immediately, and if you have lived a good and virtuous life, you may escape having already irreparably damaged the carbon canister. If it is already polluted by the excess gasoline, you will be looking at a very expensive repair that is not covered by the warranty–as owner negligence never is.

And, you have to remember that this is a repair that is unavoidable once the damage is done, as you cannot pass a state emissions test with a car in that condition. A friend of mine insisted on filling his Highlander as you do, and he wound up paying the dealer something on the order to $300.+ for replacement of the carbon canister.

For your own good, begin reading your Owner’s Manual which warns you against your method of filling the gas tank and also gives a wealth of other information that will help you to operate your car safely and economically.


I agree. In addition, the EPA testing is done with a fully warmed up engine. Typical in-town driving habits include lotsof short drives and proportionally far less mileage with the engine fully warmed up than the EPA testing does.

By the way, I find that having the windows and the roof open uses more gas on my car…because I’m having so much fun that I romp on the gas…a lot. With the windows closed I just sort of relax and cruise to my destination.


I have measured my gas mileage on a typical 360 miles round trip and on an average car (midsize); expect a drop of about 3-4 mpg with the A/C. Driving without A/C was with all windows closed on a cool day; driving with A/C was during warmer weather with all windows closed. Modern A/Cs are more efficient, but you will always have a loss.


It takes POWER to operate a A/C system. Guess where it comes from. Four cylinder engines suffer the most because the A/C demands a greater percentage of the available power. You can FEEL the compressor kick in.

By far, the BIGGEST factor in reduced mileage is acceleration. The harder and longer you accelerate, the worse mileage you will get…Drivers who get poor mileage also usually have very hot and short lived brakes…


The EPA measures exhaust emissions, not fuel consumption. Their predictions for the latter are just a byproduct of the emissions tests. The big problem is that the emissions tests are based on unrealistically conservative driving behavior. In the real world, people accelerate harder and drive faster than the EPA assumes. Since the emissions test procedure is tied to federal law, it would take, quite literally, an Act of Congress to change it.

Because the test procedure is unrealistic, the fuel consumption prediction is also unrealistic. This has been known all along. Recently, the EPA revised the way they analyze the emissions data so as to generate more realistic predictions. They are still only estimates, but should be significantly better than before. These revised predictions are shown on window stickers beginning with 2008 models. To see revised predictions for older cars, go to the EPA’s fuel consumption web site,, and click on “FIND AND COMPARE CARS”.


Mythbusters tested which gave you better or worse mileage - AC or rolling the windows down… rolling the windows down gives you worse mileage due to the drag. Just FYI for everyone :slight_smile:


You need at least 3,000 miles on the car before your gas mileage comes up a little. The drive train is still tight on a new car. You have to start the car and drive at least a half hour at a time in the city to get the city rating. You can’t get it correctly by driving two miles 56 times.