Skip to main content
Converters

MPG vs L/100km: Fuel Units That Don't Convert the Way You Expect

A fleet manager in Chicago compares two vans. The first does 14 mpg, the second 20 mpg. A fleet manager in Berlin compares the same two vans: 16.8 and 11.8 L/100km. Both managers are looking at the same trucks, and both think in "efficiency units" — yet the numbers behave differently in ways that cost real money on every fuel order.

Fuel efficiency is the one common unit category where the world split three ways, and where direct conversion hides a mathematical trap. Most units you convert daily are linear: 90 km/h is 56 mph, and every conversion table works the same in both directions. Fuel figures are not. The US and UK use miles per gallon — distance over fuel. Most of Europe, Asia, and Latin America use liters per 100 km — fuel over distance. Those are two different quantities, and the math between them is inverted.

The two world systems

  • MPG (US): miles achieved per US gallon (3.785 L). Used in the United States and loosely in some Latin American markets.
  • MPG (Imperial): miles per imperial gallon (4.546 L). Still the everyday unit in the UK, even though pumps there now sell fuel in liters.
  • L/100km: liters consumed per 100 kilometers. Standard across the EU, Canada, China, Japan's official figures, Australia, and most of the rest of the world.
  • km/L: Japan's everyday unit — liters per 100 km inverted, so unlike L/100km, higher is better.
  • MPGe: the EV/dedicated-gas metric in the US — miles the vehicle travels per 33.7 kWh of energy.

The 20% UK mistake

The most common single error: treating UK mpg and US mpg as the same number. An imperial gallon is 1.201 times larger. A car rated at 40 mpg-UK delivers about 33 mpg-US on the same fuel. Car forums are full of "my imported car's numbers don't match" threads where the answer is this one constant.

The conversion constants you need if you compute by hand:

  • US mpg to L/100km: 235.215 / mpg. So 30 mpg-US = 7.84 L/100km.
  • Imperial mpg to L/100km: 282.481 / mpg. So 30 mpg-UK = 9.42 L/100km.
  • km/L to L/100km: 100 / kmL. So 15 km/L = 6.67 L/100km.

Run 40 mpg-US through the wrong constant and you report 7.1 L/100km instead of the real 5.9 — a 20% error that propagates into fleet budgeting, import duty calculations, and comparison reviews. It is also the reason spec sheets should always state which gallon they mean, and most of them don't.

The MPG illusion

Here is the part that trips up even experienced analysts. MPG is inversely related to consumption, so equal-looking steps in mpg are not equal in fuel saved. A famous study by Larrick and Soll ("The MPG Illusion", Science, 2008) showed consumers consistently choose the wrong upgrade because of this.

Over 10,000 miles of driving per year at $3.50/gallon:

  • Going from 14 mpg to 20 mpg saves about 214 gallons — roughly $750 a year.
  • Going from 30 mpg to 40 mpg saves about 83 gallons — roughly $291 a year.

The 14→20 upgrade is 2.6x more valuable than the higher-miles-per-gallon upgrade, yet shoppers gravitate to "40 mpg looks twice as good as 20." In L/100km the same comparison is nearly impossible to miss: 16.8 → 12.4 L/100km versus 7.8 → 5.9 L/100km. The liters saved per 100 km sit right in front of you because the metric is consumption, not efficiency. Europe's unit makes the expensive decision visible; the US unit actively hides it. Keep this asymmetry in mind whenever you normalize mixed data for analysis.

Second-order trap: the labels themselves

Even after a correct unit conversion, manufacturer figures are not comparable across countries because test cycles differ:

  • EPA (US) uses a 5-cycle method that includes aggressive driving and cold starts. Its numbers run lower than European ones.
  • WLTP (EU) replaced the too-lenient NEDC in 2018 and is more realistic, but still milder than EPA combined figures.
  • Japan's WLTC city-weighted cycles skew higher (better-looking) than EU numbers for the same car.

A 5.9 L/100km WLTP figure and a 40 mpg-US EPA figure for "the same" car come from different test protocols. Converting between the units without remembering the test-cycle mismatch produces spreadsheets that look rigorous and are wrong at the category level, not the arithmetic level. Real-world fuel economy typically lands 10–25% above label figures as well, and that tolerance is not uniform across cycles.

What this matters for in practice

  • Buying a car abroad: imports into Ireland, Malta, and Cyprus come specced in mpg with UK gallons; continental listings use L/100km. Compare consumption, not labels.
  • Fleet dashboards: if your telematics system mixes mpg vehicles and L/100km vehicles, convert everything into L/100km before ranking. Ranking by per-vehicle mpg silently favors improving your best cars instead of your worst.
  • API work: open datasets state their unit in headers and docs that people skim past. Validate the column before you aggregate it.
  • Road-trip planning across a border: driving from the US into Canada, one dashboard switches from mpg and miles to liters and km/h within a day of driving.

Doing the math correctly

For occasional lookups, a dedicated fuel efficiency converter handles all five units (mpg-US, mpg-UK, km/L, L/100km, and MPGe) without you memorizing three constants. For anything batch-shaped — normalizing a fleet spreadsheet column, or reconciling telematics exports with mixed units — the batch unit converter processes the whole range in one pass. And when the trip itself crosses into the other unit world, a speed converter covers the mph–km/h side of the same confusion.

Whichever tool you use, verify the direction once on a known pair. 30 mpg-US must land on 7.84 L/100km — not 9.42 (that's UK gallons). One check number and the unit split stops being a source of silent errors.

One rule to keep

Consumption units (L/100km) tell you money saved directly: subtract, multiply by your annual distance, done. Efficiency units (mpg, km/L) require an inversion step before any savings comparison. Convert early, convert everything to the same unit family, and never rank a mixed table without doing so first.