What a Compression Test Tells You Before Buying Parts

A compression test is the measurement that tells you whether an engine's problem is worth fixing, and it is routinely misread because people look at the wrong part of the result.

A note on this guide: this rests on what the measurement can and cannot distinguish, not on engines tested here. See how we test.

The problem, as it appears

An engine down on power, using oil, misfiring on one cylinder — or simply an unknown quantity on a car you are considering buying.

Everyone suggests a compression test, and the result arrives as a set of numbers with no verdict attached: four figures that are roughly similar, or one that is clearly lower, and no clear statement of what either outcome actually means.

Consistency beats the absolute number

There is no universal "good" figure. It varies with compression ratio, engine design, even altitude, and a healthy engine of one type reads differently from a healthy engine of another.

What the test is actually good at is comparison. Four or six cylinders all reading close to each other describes an engine that has worn evenly and is behaving as a unit. One cylinder well below its neighbours is a finding — regardless of whether the absolute numbers looked impressive.

A rough guide people use is that cylinders within about ten percent of each other are consistent. The outlier is the story.

This page contains affiliate links. If you buy through them, Ugly Fixes may earn a commission at no extra cost to you. Products are selected independently.

The wet test is what turns a number into a diagnosis

This is the step that gets skipped, and it does most of the work.

Once you have a low cylinder, put a small amount of engine oil into it and take the reading again. Oil temporarily seals the gap between the piston rings and the bore.

• Reading jumps substantially — the compression was escaping past the rings. That is bore or ring wear.
• Reading barely moves — the oil could not seal the leak, so it is going somewhere else: a valve not seating, a head gasket, or a crack.

Those two outcomes have very different costs attached, which is precisely why finding out before buying parts is the point of the exercise.

Doing it so the numbers mean anything

Warm engine, then off. A cold engine reads low because clearances have not closed up.

Every plug out. With one plug in, the engine is fighting compression in that cylinder while trying to crank, so it turns slowly and every reading is depressed. The tools for getting them out without cracking one matter here.

Throttle wide open. A closed throttle starves the cylinder of air and makes a healthy engine look worn. This is the most commonly missed step.

Fuel and ignition disabled. Both. Cranking with ignition live risks a lead finding earth; cranking with injection active washes the bores with fuel. Which fuse or relay to remove is specific to your car, so look it up rather than guess.

The same number of cranking revolutions each time. Consistency in the test is what makes consistency in the results meaningful.

The mistake to avoid

Treating a low reading as a verdict without the wet retest. "Cylinder three is down" and "cylinder three is down because the rings are worn" are separated by a step that takes two minutes, and they point at completely different decisions about the car.

For a running engine that is misbehaving rather than one you are assessing, a vacuum gauge reads the whole system live and often narrows things faster. And if the car is one you are considering buying, the wider pre-purchase check covers what else is worth looking at before you get as far as removing plugs.