Fuel Economy Converter

They are reciprocals, so there is no additive conversion. Divide 235.2 by US mpg, or 282.5 by imperial mpg, to get litres per 100 km.

Because a US gallon is 3.785 litres and an imperial gallon is 4.546 — about 20% larger. The same car reads about 20% higher in imperial mpg.

Litres per 100 km is more useful for comparing savings, because it is proportional to fuel used. Improving from 20 to 25 mpg saves far more fuel than improving from 40 to 45, which mpg obscures.

The fact that equal mpg improvements do not save equal fuel. Going from 10 to 20 mpg saves five gallons per hundred miles; going from 40 to 50 saves half a gallon. It is why consumption-based units are preferred.

Official figures come from standardised test cycles that do not match real driving. Newer standards narrowed the gap without closing it, and a 10 to 20% shortfall remains normal.

Use kilowatt-hours per 100 miles or per 100 km, and compare against electricity price. Converting an EV to an mpg-equivalent requires assumptions about generation efficiency that make the number hard to interpret.

Over the journey pattern you actually drive. Hybrids do best in stop-start traffic where regenerative braking recovers energy, and their advantage largely disappears at steady motorway speeds.

The Worldwide Harmonised Light Vehicle Test Procedure, which replaced the older and more optimistic European cycle. It narrowed the gap to real-world figures without closing it.

Substantially — smooth acceleration, anticipation and steady speeds can change consumption by 20% or more on the same route in the same car.

Typically somewhere between 45 and 60 mph for most cars. Below that the engine is inefficient at low load; above it aerodynamic drag rises with the square of speed.

A few percent, more in stop-start traffic than at speed. Above about 45 mph, open windows cost more in drag than the air conditioning costs in load.

So that different vehicles are measured identically. A standardised test is comparable between cars even when it is not representative of any particular driver.