I've seen a number of sources that state you should keep your car's gas tank at least half full, some of which saying it's only important in winter. Some examples:

Cited harms include:

  • Your fuel pump overheating if it runs dry
  • Water condensation in the empty areas of the tank, freezing or causing corrosion
  • Sediment accumulating and blocking your fuel filter
  • Running out of fuel unexpectedly, leaving you in dire straits

To me, only the last of these makes some sense; the rest seem folklore-based.

So, is there any basis to these concerns?

(BTW, I tend to wait as long as possible before refueling, on the theory that I'll save time over the life of my car. I've only run out of gas a few times... ;)

  • 1
    I agree with you - "only the last of these makes some sense; the rest seem folklore-based"
    – HandyHowie
    Commented Dec 17, 2018 at 14:47
  • 3
    (Not enough for a full answer.) I run my cars usually to more than hundred thousand kilometers and have never once observed a negative effect of not filling up when half empty. I do almost the same as you: drive the car as long as reasonable before refueling. Here in the middle of Europe, even if you run out mid winter, it's no serious trouble. I've never once run out, although my range display sometimes showed all zeros for quite a bit.
    – Pavel
    Commented Dec 17, 2018 at 17:17
  • 7
    Another harm: Earthquake/hurricane requires you to get out of Dodge quickly along with 1M other people … and the gas stations are either out of gas or out of electricity to run the pumps. Keep your tank > 1/2 full so you can get down the highway fast to a useful gas station ...
    – davidbak
    Commented Dec 17, 2018 at 19:48
  • 1
    Water condensation in an empty fuel tank is a major issue for airplanes. Many accidents have been caused by water in the fuel. Pilots are required to drain our fuel and look for water before each flight. Pilots are also encouraged to keep the tanks full after each flight. However airplane fuel tanks are much larger than those in cars an more susceptible to this problem.
    – DLH
    Commented Dec 17, 2018 at 22:16
  • 2
    @davidbak Hurricanes are known days in advance. An earthquake will be over before you can get out of Dodge and driving down the road during one is probably not the best idea.
    – reirab
    Commented Dec 17, 2018 at 22:21

5 Answers 5


I'd suggest there's credence for all of these

  • Your fuel pump overheating if it runs dry

Unbeknownst to most, the fuel in the tank cools the fuel pump in most vehicles. The pump sits in a bath of fuel for just this reason. Seems counter intuitive you'd stick an electrical device into something as flammable as gasoline, but it works just fine because there's not enough oxygen in the tank to allow ignition. If the tank runs dry or near dry, heat can build up in the fuel pump which will shorten its life.

  • Water condensation in the empty areas of the tank, freezing or causing corrosion

Of the four, this is probably the least likely (at least the freezing part), though I could see it happening. In any area of the world which uses ethanol (like most of the US), the fuel blend tends to collect moisture through absorption. It does this from the air which is vented into the tank from the outside. The more air you have in your tank, the larger possibility there is something like this might happen. By keeping your tank more full, there is less space in the tank for the fuel to absorb water from, which means there's less of a chance of freezing.

Corrosion is still a factor and can be caused from excess moisture. It takes time, but it does happen. Again, keeping your tank closer to full helps preclude this issue.

EDIT NOTE: Vehicles since around the 90's have mostly been fitted with non-metal fuel tanks, so will not suffer from this like they used to. (Thanks to @FreeMan for bringing this up in the comments.)

  • Sediment accumulating and blocking your fuel filter

As per the previous portion, if you get excess moisture (which causes corrosion) the bits/pieces which flake off due to this will help to clog your filter over time.

EDIT NOTE: Along with the previous question's edit, sediment may not be prevalent from rust flaking off like it used to be with metal tanks, but sediment still does occur in the tank. Running a tank low may allow whatever sediment to exist to restrict or possibly even clog the pre-pump filter.

  • Making sure you'll never run out of fuel, leaving you in dire straits

Having grown up in the mountains of Montana where snow is a factor for four to five months out of the year, this is an extremely important thing to remember to do, for a couple of reasons.

First of all, if you should lose control of your vehicle and run off the road or get stuck for some reason, it may be a while before someone can come and rescue you. If you run out of fuel while you're waiting, you could literally freeze to death. We were taught not to run your vehicle any more than 10 minutes every hour, which should give you plenty of time if you have fuel in your tank.

Secondly, Having more fuel in your tank provides more weight to your vehicle, which helps with traction. This little extra traction can help you stay out of trouble in the first place.

  • water condensation can be an issue with diesel vehicles... And can be seen when a vehicle has a water trap combined with the fuel filter, which shows the water collected which can also be drained...
    – Solar Mike
    Commented Dec 17, 2018 at 16:07
  • @SolarMike - Good points. Commented Dec 17, 2018 at 16:09
  • 1
    @FreeMan - Concur with what you're saying here, but vehicles on the road can be older than from the 90's. I meant to have a caveat in my answer and had been thinking about it on the way home from work. I'll add it to the answer and appreciate you bringing it to light. Commented Dec 17, 2018 at 21:20
  • 2
    Note that neither sediment nor water are formed in the vapor space above the liquid in a closed tank. The water that condenses has been inside the tank before, and likewise with the sediment. Moisture gets in as humid air is sucked in as the fuel is used. Assume you have a full tank and drive until all that fuel is used up: in that time, you sucked in one tank volume of air carrying a certain amount of water. If instead you reful at 1/2, all that happens is that you suck in 2 * 1/2 tank volume of humid air: no difference at all. Similarly for particles. Commented Dec 18, 2018 at 13:26
  • 1
    @Acccumulation no, it actually is literal. If the temperature outside is below 32°F/0°C, your body temperature will slowly make its way towards freezing. During this process, you will inevitably die, and your body will continue to lose heat (faster now, even) until it is indeed frozen. So yes, you are literally freezing to death. It's just that the "death" part happens before the "freezing" part does. Now if it were 34°F out, you wouldn't literally freeze to death, you'd merely die of hypothermia without proper protection.
    – Doktor J
    Commented Dec 18, 2018 at 20:57

Your fuel pump overheating if it runs dry

This is indeed a concern, as with all other fluids in the car. Most car fluids serve dual purposes-- cooling as well as lubrication. You're not supposed to run low on anything. More volume provides more dissipation of heat.

Seals also start to dry out and crack when dehydrated. This is a concern when storing a car for a long time, which is one of the reasons why you're supposed to completely fill it with gas (and stabilizer) before storing it. This is also why the air conditioner in cars formerly owned by geriatrics tend to be broken (they never use it, so everything dries out, the pump degrades and the refrigerant escapes).

Sediment accumulating and blocking your fuel filter

I received similar advice from my dad ("never let the tank run to empty"); this is probably leftover advice from the 70s-80s when gas tanks were still made of metal and subject to corrosion. Rust flakes could clog a filter for sure.

Fuel intake is on the bottom of the tank. Like sand in your bathtub, any sediment introduced is going to naturally settle around the drain and get sucked into the filter as fuel is drawn through it regardless of volume. That's what the filter is for-- it's a consumable item. You're supposed to replace it according to your maintenance schedule anyway.

Myself, I don't live in a dusty climate so I just clean the gas cap and fuel nozzle when refueling to prevent dirt getting in in the first place.

Running out of fuel unexpectedly, leaving you in dire straits

Also a concern if you live outside a serviceable range of roadside assistance.

On the flip side, the less fuel you're carrying, the better your gas mileage (at the expense of traction).

  • 6
    I once had to tear apart a fuel tank from a rust belt '83. (City recycling didn't take fuel tanks, but did take 1' squares of sheet metal). This car had been stored for several years of its life. Anyway, the interior of the tank after 20 years was pristine. So corrosion, no. Commented Dec 18, 2018 at 1:05
  • @Harper More likely to rust from outside, especially if they're externally mounted - there was a question last week from someone with exactly that problem on an 80's VW...
    – Nick C
    Commented Dec 18, 2018 at 9:14
  • Before cars got the meters for fuel (which is something I personally only know from our ancient tractor, and from some east German cars I saw) the fuel drain was often built the way that the pipe stuck a bit into the tank. I.e. it would not drain to the last drop but instead leaves a certain volume (in German it's called Sumpf - don't know in English). In addition, they had a switch so if you ran out of fuel (which is more likely if you have to measure by stick and estimate during the drive how much you use), you could turn that switch to get some more fuel and get to the next gas station.... Commented Dec 18, 2018 at 13:32
  • ... But that fuel in the Sumpf often was more dirty because (heavy) particles accumulate in the lower layers of the fuel in the tank (unless you mix it all up by going crazy through potholes). So using the "get the last drop out" switch would get you to the gas station, but it would do this by using fuel that may be more dirty and in consequence sooner clog the filter. (Side note: I think at least here in Germany, the fuel did get much cleaner, as I don't have to change the first filter as often as we needed when I was a kid. And many cars don't have the cheap 1st filter any more at all) Commented Dec 18, 2018 at 13:36
  • @cbeleites we call it a sump, though the technical term is also unusable fuel... The particles would have to be quite heavy to not mix... Keeping laminar separation in a moving fuel tank is not possible, though, I will grant you I have never ridden fine German roads... Commented Dec 18, 2018 at 16:32

Half full seems way over the top. I normally refuel somewhere between 1/4 and shortly after the low-fuel warning coming on. If I were out in the back of beyond I would always make sure I had plenty of fuel to reach the next filling station.

I have heard (also had disputed) that if a diesel car were ever to completely run dry the resulting lack of lubrication would completely destroy the high-pressure CRDi fuel injection pump. So diesel cars are programmed to shut down and pretend to have run out of fuel while there remains a litre or two of diesel in the tank (but with the gauge below empty). Given what running even briefly on gasoline does to an injector pump, I can certainly believe this.

A very low fuel level may combine with vehicle motion to swirl any sediment out of the tank and into the fuel line, so a filter blockage is also a plausible concern.

But both these concern running on the dregs, not running below half full.

BTW there is a small but non-zero economy penalty for filling up early. The average weight of fuel you are lugging around will be higher, so your fuel economy will be worsened.

  • 1
    (Diesel) injection pumps rely on the lubrication by the diesel. Not only the high pressure ones, but the old low pressure ones as well. Whether they are completely destroyed by a single event of running out of fuel is IMHO another question, particularly as I'd expect running out of fuel stops the engine before the injection pump is dry: as soon as a bit of air gets into the pressure side of the fuel system, the injection pump works against a compressing/expanding gas bubble and no further fuel is transported. Even with the old low pressure diesels, running air in the injection system meant.. Commented Dec 18, 2018 at 13:51
  • 1
    ... a lot of hassle until all air was out (btw, many of these did not have an electric fuel pump but also relied on bubble-free filling on the tank side: that allowed sucking in the needed fuel): first (manually) pumping until the low pressure side was bubble-free. Then one cylinder after the other loosening the high pressure pipes and having the injection pump pumping until no further bubbles occured. Fasten the pipe, try to start. Repeat until no bubbles left and engine running... That was with compared to nowadays low pressure tech that you could do if needed without a workshop as long... Commented Dec 18, 2018 at 13:56
  • 1
    you had the right wrenches. I didn't have the occasion so far to de-air common rail engines, but I think already the saved hassle is for sure worth shutting off the engine before any air gets into the high-pressure side - and if that's an additional safety measure for the injection pump, fine. Commented Dec 18, 2018 at 14:00
  • 1
    The feed pump [Förderpumpe - not sure about some transations here] is a small electric pump that will work to transport fuel as soon as fuel is available again, so problem there (that's the pump I said was absent in many of the old diesels). The injection pump is AFAIK a piston pump, and it doesn't matter how it is driven (mechanically or electrically): if the piston moves in a space completely filled with incompressible liquid, that liquid has to go somewhere and the pressure builds up accordingly. If there's an air bubble, that acts as a pressure buffer. As the air bubble exceeds a certain Commented Dec 18, 2018 at 14:08
  • 1
    volume, the pump doesn't reach the pressure to fill the high pressure reservoir = common rail at the required pressure (old diesel: open the nozzle - same for high pressure systems, but they do have sensors that would shut off the engine already at far higher pressures)/open the outlet valve. Plus for the high pressure systems, we're in a pressure range where AFAIK the compressibility of the fuel has to be taken into account as well. Now, larger piston replacement volume compared to the dead volume will still work with larger air bubble. But there are technical limits/other drawbacks for large Commented Dec 18, 2018 at 15:03

I used to own a car that ran on LPG (Honda Accord). This meant my LPG tank was my main fuel source, and the full-size petrol tank was used only when starting the engine, and as a backup. I tended to put about 20 litres of petrol in every 3 months, so the tank was never more than 2/5 full. When I sold the car it was 12 years old and ran without issue, suggesting points 2 and 3 are not major factors. I live in .nl where the climate is temperate and fuel is of high quality, which would influence those points.


With all the good input, and with a bit more research, I'd like to answer my own question. So, here were the concerns I raised on letting a car's fuel tank get close to empty:

Concern 1: your fuel pump would overheat if it ran dry

In-tank fuel pumps are cooled by the flow of gas around the motor:

enter image description here
Source: www.aa1car.com

So, there is a valid concern that running the pump dry will leave the motor uncooled, causing damage. However, pumps seem to draw less than 10 amps (see here, here, and here) which at 12 volts is probably less than 100 watts. That isn't much power; it would take at least fifteen minutes for it to heat a cup of water to boiling. You probably shouldn't continually try to crank a car with an empty gas tank, but I doubt a second or two of fuel-free running would overheat a pump.

Concern 2: water would condense in the empty areas of the tank, freezing or causing corrosion

Water can get into your fuel tank in two ways: mixed into the pumped gas when filling, or as humid air being pulled in as fuel is being pulled out and used. For the former, the amount of water is just how much gas you pump; it doesn't matter how you split it up into fillings. For the latter, as cbeleites noted, emptying a tank halfway twice pulls in exactly as much air as emptying a tank completely once, so again you'll get the same amount of water pulled in no matter how you split that up into fillings.

As freeman noted, OEM fuel tanks these days are plastic, not metal, so corrosion isn't going to be an issue. In terms of freezing, if there's ethanol in the gas then the water will likely blend with the gas, so it won't freeze. Any water that doesn't mix in will sink to the bottom and probably get pulled through the fuel pump. This could be a problem, but the magnitude of the problem won't depend on whether or not there's lots of gas floating on top of the water.

Concern 3: sediment would accumulate and block your fuel filter

This doesn't seem likely to me, assuming you buy from semi-reputable gas stations. But, if a gas station does sell you dirty gas, or if someone spoons mud into your gas inlet, it would sink to the bottom and likely be sucked into the pump whether or not there's lots of gas in the tank. Yes, the sloshing of a less-than-full tank would move the sediment around and make it more likely to get pulled into the pump, but unless you drive a limo on really smooth roads, or keep your tank almost completely full, that's still going to happen. And, as ivan noted, that's what fuel filters for; they are designed to capture silt and be replaced at regular intervals.

Concern 4: you might run out of fuel unexpectedly, leaving you in dire straits

This is a real concern, but it has to be evaluated based on where you live and what your priorities are. I live in the New England suburbs, where there's a gas station every few miles, and even on the highway I'm clear about how far I need to go and how far my tank will take me. And, if I make a mistake and do run out of gas, it will be inconvenient, not dire. If you want to make sure you can always drive a couple of hundred miles at a moment's notice, then you should keep some gas in your tank. If not, don't worry about it.

Postscript: the benefit of filling only when your tank is empty

My car gets about 25MPG, and has a 14-gallon tank. If I drive it 120k miles then that's 4800 gallons of gas over its lifetime. If I fill my tank up completely each time, that's 340 fills; if I fill it halfway then that's 680 fills.

If it takes me five minutes to get to and from the gas station (ignoring the time spent actually standing at the pump), then 340 fills would take about 28 hours, and 640 fills would take about 56 hours. So, by only refilling when empty I'd save about 28 hours over the life of my car. Seems worthwhile to me.

Thanks for everyone's input! Additional feedback welcome!

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .