How to calculate fuel mileage: the method, the math, and the gotchas

The formula

Fuel mileage, in any unit, is one equation:

That is the whole calculation. The job is making the inputs accurate.

Why a single fill-up is noisy

One fill-up can vary by 3 to 5 MPG from the next, even with the same driver on the same route. The variation comes from cold-start fuel enrichment, the small differences in pump calibration between stations, the difference between filling to the first click and topping off, and the small lump in your arm that adds 0.2 gallons of fuel to the "tank is full" reading. None of these are mistakes; they are the noise floor of the measurement.

Three to five fill-ups, treated as a single sample, reduces the noise to something you can compare to the EPA combined rating. Ten or more, calculated the same way each time, is what the SPC guide uses to detect a real change in efficiency (an aging sensor, a dragging brake) rather than just noise.

Step-by-step

The fill-up method, in seven steps. The complete walkthrough with troubleshooting is on the step-by-step guide; this section is the short version.

1. Fill the tank to the first click. Do not top off. The first click is the most repeatable fill level.
2. Record the odometer to the nearest 0.1 mile. Photograph it. If your car has a trip meter, zero it; if not, write the odometer down.
3. Drive normally until you have used at least half a tank. Aim for 200 to 300 miles between fill-ups. Below 200 miles, measurement noise dominates.
4. At the next fill-up, use the same pump at the same station. Pump-to-pump calibration can vary by 0.5 to 1.0 percent.
5. Fill to the first click again. Note the gallons to 0.01. The pump display is more accurate than the receipt total.
6. Record the new odometer reading. Photograph it.
7. Divide miles driven by gallons used. That is your MPG for the cycle.

The full procedure, including what to do when you cannot start with a full tank, how to handle partial fills, and the photo-validation habit that catches more errors than any other single change, is on the step-by-step guide.

Metric and other units

Outside the US, the same calculation is reported as L/100km, which is the inverse: lower is better. The two are not linearly related. See the conversion guide for the math and the "reciprocal trap" that catches people who assume a 20 percent drop in L/100km equals a 20 percent rise in MPG.

The constant 235.214583 comes from 100 km ÷ 1.609344 miles × 3.785412 L/gallon, the same conversion used by every major automaker and by fuel-economy regulators worldwide. The UK and the US use different gallons (4.546 L vs 3.785 L), which is why a "40 MPG" UK figure is a "33 MPG" US figure for the same car.

For worked conversions across vehicle classes, see the MPG vs L/100km conversion guide; the math derivation is on the MPG formula math page.

Averaging across fill-ups

For a vehicle you track over time, the question is "what is the average MPG for this car, this driver, these conditions." Two methods give different answers.

Simple average (don't do this)

Add up the MPG from each fill-up, divide by the number of fill-ups. This underweights long fill-ups and overweights short ones. If you have a 250-mile fill-up at 28 MPG and a 50-mile fill-up at 22 MPG, the simple average is 25 MPG, which is misleading; you actually burned 8.93 gallons in 250 miles and 2.27 gallons in 50 miles, for 300 miles on 11.2 gallons, which is 26.8 MPG.

Weighted (total-distance) average (do this)

Add the total miles across all fill-ups, add the total gallons, divide. This is the right answer:

For a 3 to 5 fill-up sample, this is the number to compare to the EPA combined rating. For a 10+ fill-up sample treated as a process (the SPC guide approach), this is the centerline of your control chart.

Common mistakes

Most "my MPG is way off" complaints trace to one of these:

1. Not filling to the same level each time. If you fill to the first click on fill 1 and top off on fill 2, the second fill-up is short by the topped-off amount (0.5 to 1.5 gallons is typical). Always stop at the first click.
2. Reading the odometer wrong. 45,230.2 and 45,230.5 look the same in the rain. Photograph the odometer at each fill-up. This catches more errors than any other single habit.
3. Mixing units. Odometer in miles, fuel in liters, or vice versa, without converting. The conversion constants are on the math page, but the safer fix is to use the calculator, which handles them for you.
4. Calculating on a single fill-up. One tank is noise. Three to five fill-ups is a sample. Ten or more is a defensible measurement.
5. Comparing to the EPA sticker without context. EPA ratings are lab-tested under controlled conditions. Real-world MPG is typically 10 to 15 percent lower. The variation comes from weather, traffic, A/C use, fuel blend, and your driving. The factors guide breaks these down.
6. Resetting the trip meter between fill-ups and forgetting. A partial tank of fuel gets measured into the next fill-up. If you reset, that fill-up is uncomputable.

What the calculation does not tell you

The number you get from this method is an average over the fill-up cycle. It does not tell you:

• Instantaneous MPG. For per-second fuel-use data, you need an OBD-II adapter. The OBD-II guide covers what consumer hardware can and can't do.
• Specific driving-condition MPG. To compare city to highway, you need to log the type of driving in each cycle and split the average. The EPA does this on a dynamometer; on the road, it requires discipline.
• Why your MPG is what it is. The calculation gives you the number; it does not diagnose the cause of a low number. Mechanical causes (dragging brake, low tire pressure, O2 sensor) and driving-style causes (aggressive acceleration, excessive idling) are different fixes. The factors guide and the accuracy optimization guide cover both.
• An authoritative figure for warranty or tax purposes. The calculator's result is an estimate, not a measurement. For an authoritative figure, use the manufacturer's published number or get a controlled test.

Frequently asked questions