Compression
Compression testing is the single most valuable mechanical health check you can run on an engine without pulling it apart.
- A wet test (small amount of oil in the cylinder) will raise compression if the leak is at the rings — oil temporarily seals worn rings. If wet numbers match dry, the leak is at valves or head gasket, not rings. This is the fastest way to narrow root cause.
- Air escaping into the crankcase (heard at the oil filler) indicates ring leakage. Intake valve leakage would be heard at the throttle body, exhaust at the tailpipe, and head gasket at the coolant reservoir or an adjacent cylinder.
- Industry standard: no cylinder should differ from the highest by more than 10%. Beyond that, drivability and emissions issues typically appear. Some manufacturers spec tighter (7-8%).
- Air at the tailpipe means the exhaust valve isn't sealing. Common causes: burnt valve, valve seat recession, carbon holding valve open, or bent valve from timing failure. A tear-down is required for repair.
- Compression tests must be done at operating temperature (rings and bores at proper clearances), throttle wide open (unrestricted airflow), and with all plugs removed (no drag on adjacent cylinders). Cold tests read low across the board and are misleading.
- Diesel engines have very high compression (350-500 psi typical). Low readings indicate the same causes as gasoline (rings, valves, gasket) but the higher pressures require a diesel-rated compression tester (installed via glow plug or injector hole) — not a gasoline gauge.
- Compression is only one leg of the combustion triangle. Missing spark (bad coil or plug), missing fuel (bad injector), or extra air (vacuum leak at intake runner) all cause misfire with good compression. Systematic isolation identifies which one — swap parts to good cylinder, observe change.
- Oil in the cylinder temporarily seals gaps at worn rings or cylinder walls, allowing compression to rise. Valves seal against the head and aren't affected by cylinder-wall oil. Big rise on wet test = ring wear. No change or minimal change = valve/gasket issue.
- Good compression rules out mechanical wear as the primary cause. Progressive power loss usually means restriction (intake or exhaust) or fuel delivery limitations under load. Test with scan-tool data at wide-open-throttle: fuel trims, MAF readings, boost, and back-pressure measurements narrow the cause.
Every combustion event depends on the cylinder sealing pressure between the rings, the valves, and the head gasket, so a compression or leakdown test tells you where the seal is failing before you commit to a tear-down. Learn to run these tests correctly and read them precisely, and you will stop guessing and start diagnosing.
Running the Test the Right Way
A compression test is only as good as the conditions you run it under. The professional standard is to test at operating temperature, because the rings, pistons, and bores only reach their designed clearances when the engine is hot. A cold engine reads low across every cylinder, and a tech who condemns an engine off a cold test has misled the customer. Hold the throttle wide open so each cylinder can pull a full charge of air, and remove all the spark plugs so the other cylinders are not dragging down cranking speed. Skip any one of these and the numbers lose their meaning.
On a healthy gasoline engine, what matters most is not the absolute number but the spread between cylinders. The industry standard is that no cylinder should read more than 10 percent below the highest cylinder; some manufacturers spec even tighter at 7 to 8 percent. Beyond that spread, you can expect drivability complaints and emissions problems.
Diesels play by different rules. They rely on compression heat to ignite fuel, so cranking pressures typically run 350 to 500 psi. A diesel reading of 300 psi where 400-plus is expected points to the same suspects as a gas engine, worn rings, worn cylinder walls, or valve sealing problems, but before you condemn anything, verify your equipment. Diesel compression testing requires a diesel-rated gauge installed through the glow plug or injector hole; a gasoline gauge is not built for those pressures and will lie to you or fail.
Wet Testing: Rings or Valves?
When one cylinder reads low, say 90 psi while its neighbors sit at 165, the next move is not a rebuild quote and not a head gasket. It is a wet compression test. Squirt a small amount of oil into the low cylinder and retest. The oil temporarily seals the gaps at worn rings and worn cylinder walls, so if the leak path is at the rings, the compression number rises significantly with oil in the bore. Take the oil away and the low number comes right back.
Here is the logic behind it: valves seal against seats machined into the head, up and away from the cylinder walls, so oil pooled on the piston crown does nothing for a leaking valve. A big rise on the wet test means ring or cylinder wall wear. No change, or only a minimal change, means the leak is at a valve or the head gasket. This one simple test is the fastest way to split the diagnosis in half and decide whether you are looking at a bottom-end job or a head job.
Leakdown Testing: Following the Air
A leakdown test pressurizes the cylinder with shop air at TDC compression and measures the percentage that escapes. Its real power is that the escaping air tells you exactly where it is going. Get in the habit of listening at four places.
Air hissing from the oil filler cap means combustion pressure is blowing past the rings into the crankcase, so a reading like 25 percent leakage heard at the filler is a ring or cylinder wall problem. Air at the throttle body means an intake valve is not sealing. Air at the tailpipe means the exhaust valve is not seating; common causes there include a burnt valve, valve seat recession, carbon holding the valve off its seat, or a valve bent by a timing failure, and every one of those requires tear-down to repair. Bubbles in the coolant reservoir, or air crossing into an adjacent cylinder, point to head gasket failure.
Leakdown also confirms an inter-cylinder gasket failure. If cylinders 3 and 4 both test low on compression, pressurize one of them and listen or watch for air escaping into its neighbor. Air crossing between adjacent cylinders proves the gasket has failed between them.
When Compression Is Good but the Engine Still Struggles
Do not tunnel-vision on compression. A cylinder can hold perfect pressure and still misfire, because compression is only one leg of the combustion triangle. If a cylinder with normal compression misfires, work the other legs: ignition (a failing coil, plug, or wire), fuel (a dead or dirty injector), or air (a vacuum leak at that cylinder's intake runner). The systematic move is to swap the suspect component to a known-good cylinder and see whether the misfire follows it.
The same discipline applies to an engine losing power steadily under load with all cylinders at spec. Good compression rules out mechanical wear as the primary cause, so look for what limits the engine under demand: fuel delivery under load (fuel pressure and injector capacity), timing drift from a stretched belt or chain, intake restriction (a dirty air filter or a boost leak on a turbocharged engine), or exhaust restriction from a clogged catalytic converter or muffler. Scan-tool data captured at wide-open throttle, fuel trims, MAF readings, boost pressure, and exhaust back-pressure measurements, will narrow it down without guesswork.
| Measurement | Healthy | Interpretation |
|---|---|---|
| Gasoline cranking compression | Typically 125–180 psi | Absolute value tracks compression ratio — the spread matters most |
| Cylinder-to-cylinder spread | Within 10% of the highest | Beyond = misfire, drivability, and emissions trouble |
| Diesel cranking compression | ~350–500 psi | Requires a diesel-rated gauge through glow plug or injector hole |
| Wet test response | Large rise = rings/walls | Little or no change = valve or head gasket |
| Leakdown, healthy | Under ~10% (fresh engine ~5%) | Where the air exits names the leak path |
| Leakdown, action level | 20% and up | Locate escape: filler cap, intake, tailpipe, coolant |
| Running compression at idle | Roughly 50–75% of cranking value; snap ~80% | Exposes dynamic sealing and breathing problems a cranking test hides |
Always test warm, throttle wide open, all plugs out, with healthy cranking speed. Verify against the engine's published spec.
- Warm engine, all plugs out, throttle blocked wide open, fuel and spark disabled. Crank each cylinder the same count of compression pulses — four to six — and record every dry number.
- Read the pattern before reaching for tools: one low hole is a local problem; two adjacent low holes suspect the gasket between them; all cylinders evenly low points at timing, cam wear, or uniform ring wear.
- Wet-test only the low cylinders: a teaspoon of oil in the bore, retest. A big rise means the oil temporarily sealed worn rings or walls; little change clears the rings and indicts a valve or the gasket.
- Leakdown the low cylinder at TDC compression and listen: air at the oil filler cap is rings, at the throttle body is an intake valve, at the tailpipe is an exhaust valve, and bubbles in the coolant reservoir mean the gasket or a crack.
- For an adjacent low pair, pressurize one cylinder and listen in its neighbor — air crossing between them proves the gasket has failed between the bores.
- Quote from the evidence. Which seal failed — rings, valve, or gasket — is the difference between a bottom-end job, a head job, and a gasket job, and now you know instead of guessing.
- Compression testing cold and condemning the engine. Rings and bores only reach designed clearances hot — a cold test reads low across the board and turns a healthy engine into a rebuild quote.
- Testing with the throttle closed. The cylinders can't pull a full air charge, every reading comes back low and even, and the pattern mimics a uniformly worn engine.
- Quoting a rebuild off one low dry reading. The wet test and leakdown split rings from valves from gasket in twenty minutes — the difference between a $1,500 head job and a $6,000 rebuild deserves twenty minutes.
- Chasing compression on a misfiring cylinder that tested healthy. Compression is one leg of the triangle — a cylinder holding perfect pressure still needs spark, fuel, and air, so move the diagnosis to the other legs instead of re-testing the one that passed.
Cylinder 3 reads 95 psi dry while the others hold 160. Wet, it jumps to 150. Verdict?
Rings or cylinder wall wear in cylinder 3. The oil temporarily sealed the gap at the rings, which is why the number recovered — a leaking valve sits up in the head where pooled oil can't help it. Confirm with leakdown expecting air at the oil filler cap, then the conversation is honing and rings versus boring, not valves.
Cylinders 2 and 3 both read 90 psi; the rest read 165. Most likely single cause, and the confirming test?
Head gasket failure between the two adjacent cylinders — one breach explains both low holes. Confirm by pressurizing cylinder 2 with the leakdown tester and listening at cylinder 3's plug hole: air crossing between them proves the bridge is breached.
Every cylinder reads an even 100–110 on a high-mileage engine that runs but has no power. What causes even-low across the board?
Uniform causes: worn rings across all cylinders on a tired engine, retarded cam timing from a stretched chain or jumped belt, or a flawed test — cold engine, closed throttle, or slow cranking from a weak battery. Verify the test conditions first, then check cam/crank correlation on the scan tool before opening anything.
Cylinder 3 reads 95 psi dry while the others hold 160. Wet, it jumps to 150. Verdict?
Missed one? The reasoning above comes straight from the ENG exam bank, so this is the standard you will be held to.