Suppose a very hot summer day, driving on a long straight highway at 130 km/h. What is most fuel efficient solution :

  • Driving with AC on and closed windows?
  • Driving with AC off and four full opened windows?

One uses fuel for running AC, other uses fuel from bad aerodynamics due to open windows.

This may vary depending on car model and AC model, yet is there a way to know which case is most fuel efficient in a general way?

  • 14
    The third option is the keep the windows up, and have the blower on but no AC, this is inevitably the most efficient. – Phil G Jan 14 '20 at 15:22
  • 7
    Also done by mythbusters. – JPhi1618 Jan 14 '20 at 21:48
  • 7
    Are you sure you want to drive at 130km/h with four fully opened windows? I think this would require full motor cycle gear to feel comfortable. At say 30km/h sure, all windows wide open, but at 130km/h even two windows open just a few centimeters provides quite a lot of cooling. – quarague Jan 15 '20 at 7:45
  • 4
    Answer to title question: A glider. Freefall also works, but for more limited duration. – reirab Jan 15 '20 at 16:04
  • 4
    Does this answer your question? Is it more efficient to drive with the windows down or the AC on? – cawwot Jan 15 '20 at 19:11

It depends on several factors:

  • The car's aerodynamics
  • The AC compressor load
  • Temperature differential between the desired temp and the outside temperature

The SAE did an experiment, the full results are paywalled, but there's some useful detail in the summary:

On-road and laboratory experiments with a 2009 Ford Explorer and a 2009 Toyota Corolla were conducted to assess the fuel consumption penalty associated with air conditioner (A/C) use at idle and highway cruise conditions. Vehicle data were acquired on-road and on a chassis dynamometer. Data were gathered for various A/C settings and with the A/C off and the windows open. At steady speeds between 64.4 and 113 kph (40 and 70 mph), both vehicles consumed more fuel with the A/C on at maximum cooling load (compressor at 100% duty cycle) than when driving with the windows down. The Explorer maintained this trend beyond 113 kph (70 mph), while the Corolla fuel consumption with the windows down matched that of running the A/C at 121 kph (75 mph), and exceeded it at 129 kph (80 mph). The incremental fuel consumption rate penalty due to air conditioner use was nearly constant with a slight trend of increasing consumption with increasing vehicle (and compressor) speed. A lower fuel penalty due to A/C operation is observed at idle for both vehicles, likely due to the low compressor speed at this operating point, although the percentage increase due to A/C use is highest at idle.

So at maximum AC it was more efficient to have the windows open at lower speeds, and it differed at higher speeds depending on the vehicle. At a lower compressor load the trade-off speed will be lower, again depending on the vehicle.

Mythbusters did an episode which replicated this finding, but there were some issues in the test method which make the result not quite scientific enough to use.

So the short answer is it's probably more efficient to have the windows down, unless you are going pretty fast in an aerodynamic vehicle. There are many factors in this, it's impossible to say without testing in specific vehicles.

  • 5
    But most ac use is not at 100% - they match the compressor to the load and, the ecu in most cars now will cut the ac when you accelerate hard so the engine power gets to the wheels... and the inherent thermal capacity of the ac system catches up when the compressor comes back in. – Solar Mike Jan 14 '20 at 17:55
  • 3
    @SolarMike, all that only changes the crossover point. The energy draw of the AC is constant, the energy draw of the windows increases with the square of speed, and at some point, the two curves will intersect. – Mark Jan 14 '20 at 23:53
  • 4
    The occupants of the vehicle are likely to adjust how open the windows are based on the speed of the vehicle. There's quite a bit of difference between what the interior of the vehicle experiences with wide open windows at 64.4kph (40mph) vs 129kph (80mph). While, the comparison of max AC to fully open windows does give some helpful data-points, it doesn't really tell us the relative cost of maintaining a specified comfort level, which is probably different for the two cooling methods. Perhaps there's more/better data behind the paywall. – Makyen Jan 15 '20 at 2:12
  • 2
    Regardless of being more efficient or not, the noise that the wind makes will be deafening. Possibly literally. Just close the windows and turn the A/C on, in my opinion. – Ismael Miguel Jan 15 '20 at 9:11
  • 3
    @IsmaelMiguel: yeah, since it's at least close to break-even even at lower speeds, you can easily justify the big advantage in comfort. Of course, if you care that much about fuel efficiency, you would normally avoid driving 130km/h at all, unless there's enough traffic at that speed that the only safe option is going with the flow. Speeds like 90 km/h are more efficient and not too much slower (than the 100 or 110 speed limits in Canada). – Peter Cordes Jan 16 '20 at 2:35

Keeping a constant speed using cruise control with windows shut and ac on because once the cabin is down to temperature it requires little to keep it there and the ac will modulate its power needs.

Driving with windows open will increase the drag and that will only reduce if you either reduce speed or close the windows.

  • 4
    Not only it increases drag, it makes an horrible loud noise. – Ismael Miguel Jan 15 '20 at 9:14
  • 1
    Each human in the cabin provides about 400W of heating - especially if there are many people in the car that sums up. – Nobody Jan 16 '20 at 9:24
  • @Nobody and what are the losses through doors windows and roof when the outside temperature is -20 deg C? Also the wind chill? – Solar Mike Jan 16 '20 at 9:34
  • I thought the question was about a hot summer scenario. – Nobody Jan 16 '20 at 9:37
  • 2
    @Nobody1 More like 100W per passanger unless they're exercising in the car. – nitzel Jan 16 '20 at 11:07

Fuel efficiency shouldn't be your only consideration. Rolled down windows will expose you to noise levels well above 85 decibels which can cause permanent hearing damage with prolonged exposure.

British scientists tested everything from a Mazda to an Aston Martin in non-rush hour traffic going 50, 60 and 70 MPH.

They found that regardless of speed or model, when the windows were down the driver was exposed to a nearly constant 89 decibels of noise.


  • That 1. is something that will be obvious to anybody trying it out, 2. isn't the only reason driving with windows down is uncomfortable, 3. can be easily mitigated with ear protection, 4. is a frame challenge to the question asked, and should thus be a comment, not an answer. Therefore -1 from me. – leftaroundabout Jan 16 '20 at 13:42
  • @leftaroundabout 1. People will drive miles to save pennies on a gallon of gas. Even if it's obvious, I have minimal trust in common sense. As a father, I fear for children that get subjected to this abuse. 2. Agreed, see #1. 3. Better check with local laws before obstructing your hearing while driving. drivinglaws.aaa.com/tag/headsets and thenewswheel.com/… 4. You are very much free to vote as you please but my answer is most certainly not a mere comment. – MonkeyZeus Jan 16 '20 at 13:53
  • “fear for children”: that's a valid point. — “check with local laws”: indeed you should, but good hearing protection doesn't really “obstruct” your hearing. In fact, one is usually able to understand sounds better when wearing protection, because the ears are less overloaded by the noise level. (Though, again you might validly remark that people would be more likely to just stuff in cheap earplugs rather than quality form-molded ones, and those foam things do significantly obstruct hearing). Apart from that: the open windows facilitate hearing outside significantly better to begin with. – leftaroundabout Jan 16 '20 at 16:18
  • @leftaroundabout It's amazing how many problems you create for yourself for wanting to roll you windows down. I don't think it is possible for the minuscule gas savings (if any) to ever outweigh the cons. One more thing, you introduce a lot more dirt into your car with the windows down. – MonkeyZeus Jan 16 '20 at 19:07

Suppose a very hot summer day, driving on a long straight highway at 130 km/h. What is most fuel efficient solution :

  • Driving with AC on and closed windows?
  • Driving with AC off and four full opened windows?

Of several choices, those are the worst. Why are you confining your selection to those? Also, running windows-open will make you deaf.

One uses fuel for running AC, other uses fuel from bad aerodynamics due to open windows.

That is why they are both lousy choices. Here is your better choice.

  • Windows closed, A/C off, fan on max.

We are at the best choice available given the speed you want to go. This should improve your fuel economy significantly over the other two.

Your best choice, however, is

  • Slow down.

It matters more than you think. At 130 kph, the vast majority of your fuel is being spent to overcome aerodynamic drag.

  • This is a really bad answer, a fan is absolutely not a solution if it is even reasonably hot and lowering your speed is not an option if that's the speed everyone else is traveling at (unless you want to be a danger to everyone else on the road). – eps Jan 16 '20 at 18:38
  • @eps I've driven for over 20 years exactly that way. You don't get any hotter than you would with the windows open. (in other words, when it's miserable hot, opening the windows does not improve things). And 130 kph is over 80 mph, precious few places where that's actually the flow of traffic. Nowhere in America (even when the road is carded 80, actual flow is 70-75. Really. Because 95% of the traffic is semi's.) And I am a speed demon who does drive 85 where carded 80. And I do it a lot. – Harper - Reinstate Monica Jan 16 '20 at 18:47

We could make a quantitative comparison using some ballpark figures.

Power consumption of your car's air conditioner depends on many factors, including its settings and the environment. A fair assumption of average consumption is approximately 3 kW (which might be as much as a factor of 2 off). This number can be compared with The increase of air drag power losses by opening windows.

Air drag power losses can approximately be calculated as P = ½ ρ C A v3, with ρ the air density, C the vehicle air drag coefficient, A the vehicle frontal surface area and v the vehicle velocity. We have ρ = 1.225 k/m³ and v = 130/3.6 km/h = 36.111 m/s. A and C largely depend on the car type. For a modern car (sedan) we have C A ≈ 0.3×1.8 = 0.54 and for an MPV/SUV C·A ≈ 0.45×3 = 1.35. We find P = 15.7 kW and P = 38.9 kW for respectively the sedan and MPV. The next question is: how do these figures increase if windows are opened? A study has shown that power losses increase with ca. 8% and 20% for sedans and SUVs respectively. Notably, this difference in percent seems to scale inversely linear to C A, suggesting that the increased losses are not so much a function of the car's shape. This makes sense: car windows have approximately the same size in both big and small cars, and opening them does not change your car's frontal shape. It follows that the increased losses can be roughly described by Δ P = ½ ρ 0.108 v3, for which we find 3.1 kW (for 130 km/h), approximately the same as the 3 kW figure of the air conditioner!

Be reminded that various of the above figures are very rough, including the 3kW figure of the air conditioner. As such, we cannot come to an obvious conclusion. However, also note that air drag increases cubically with speed. It will likely be safe to say that, for speeds under 100 km/h (ca. halve the drag) open windows will win, and for speed above 160 km/h (ca. double the drag) the air conditioner will win. Anyway, as others have mentioned, for higher speeds (>80 km/h), other issues will become more prominent, such as noise, for which closed windows might nevertheless be preferred.


If you want to save fuel, forget window vs. AC, the obvious solution is: drive 100km/h rather than 130km/h.

Although the difference seems like not that much (merely going 25% slower), you need 2.2 times as much energy to drive 130 instead of 100 km/h. Velocity is the single biggest deciding factor in the equation, as it's raised to the third power. Anything raised to the third power isn't very interesting with small numbers, but becomes crazy as numbers get big.

Not the answer you're looking for? Browse other questions tagged or ask your own question.