Is Leak Down Test outdated?

Thanks for your participation for earlier threads.

As I talk with with some folks of my posted issues, I am learning about relevancy of Leak Down Test for recent modernized auto engines.

What are the engine diagnostics that supersedes Leak Down Test?

Are there other tests respond better in auto engine diagnostics?

Thanks for sharing.

A dry compression test…followed by a wet compression test would normally precede a leak down test.

As far as other test’s…it all depends on where you think your problem might be.


The basic mechanical design of the four-cycle engine is the same as it was 50 years ago. The difference today is in the metallurgy and machine tolerances and the much more precise electronic engine management, which includes fuel injection and ignition systems. So, basic diagnostic procedures are still very viable to determine the soundness of the mechanicals. Compression testing and leak-down testing is still done, as is vacuum testing. I don’t know of any other testing that supersedes any of it. ODB-II trouble code diagnostics only get you so far.

The purpose of a leakdown test is to determine if the cylinder(s) are sealing properly when both/all valves are closed and, if a cylinder is not, identify it for further analysis. If a cylinder is found not to be sealing AND it’s in unison with a test of the coolant that shows hydrocarbons in the coolant, it’s’ a safe bet that a headgasket leak is in evidence and removal of the cylinder head is in order. If not, there is a possibility that a valve isn’t seating properly for some reason, and removal of the valvecover might provide additional information… however it’s safe to assume that whether it’s a valve or a headgasket, you’re going to need to remove the head anyway for a final diagnosis and repair.

Leakdown tests are not only for modern engines. All 4-cycle internal combustion reciprocating engines should seal the chambers when both/all the valves are closed, no matter the age or size. There are allowable leakages, but if you have a problem you’ll know it; just as a balloon with a pinhole won’t hold air, neither will a cylinder with a leak.

Your comments suggest that while you’re certainly learning, you don’t yet have a firm grasp on exactly how a reciprocating engine works. Having that knowledge helps make many or the things we write (and argue about) understandable and clear. If you’re interested, the following link provides what I believe to be an excellent primer on the subject. I recommend a visit.

Regarding the question of other tests being better, it doesn’t work on an ascending order; different tests provide different information to a diagnostician. Putting the ignition system on an oscilloscope cannot diagnose a leaky valve any more than doing a cylinder leakdown test can diagnose a bad coil. Every test has its purpose, its strengths, and its weaknesses. If an engine is overheating, for example, there are a number of tests a diagnostician can do to find the cause, depending on what the overall symptomology is and depending on eth results of prior tests. If a cylinder is misfiring, there are a number of possible causes that the diagnostician can test for. Yet in diagnosing either the diagnostician can end up finding a headgasket leak… or something entirely unrelated.

I hope this helps.

I agree with Yosemite and a compression test has always pretty much told me what I need to know.

A precursor to a compression test might be a vacuum gauge test. A vacuum test will not reveal a problem with a specific cylinder but can reveal a lot about engine wear, vacuum leaks, weak or broken valve springs, etc, etc. The vacuum test generally only takes a couple of minutes anyway.

Thanks for your suggestion.

I learned that engine should be in running condition to get correct result through Leak Down Test.

If we have a broken timingbelt situation, the cam timing will be off.

How does Leak Down test respond in this situation?

Does manually rotating the cam to do Leak Down Test provide accurate result?

If not, what are the tests performed in this situation to assess Engine Health?

Thanks for sharing.

With a broken or jumped timing belt a leakdown test is perfectly valid as long as the camshaft (or plural) is manually rotated to assure that all of the cylinder head valves are closed on the cylinder being tested.

A compression or vacuum test is not valid on an engine with a broken or jumped timing belt. The must be redone in order to perform those tests and in regards to the vacuum test the engine must be running.

For those with a talent for things mechanical and a few years of experience working on engines tests are often used to confirm their hunches. For me a leak down test was somewhat common on push rod engines when it was easy to loosen rocker arms to allow valves to close. All the gear heads here will understand how that test would be usefull on 400 small block Chevy engines. However, rotating a camshaft on OHC engines with the timing belt off can result in bending valves when air pressure forces a pair of pistons down which results in two pistons being forced up to top dead center possibly against open valves. But it’s difficult to second guess what other mechanics might have done based on second hand information from the car owner who is not familiar with the situation.

Often mechanics are so certain that valves are bent when they find a bad timing belt that they choose not to install a new timing belt and test it but rather look to the quickest method to substantiate their “hunch.” But I don’t know how experienced the OP’s mechanic is with the specific engine. The man might have already attempted to start a dozen of those engines after replacing the belt and all of them failed to start due to bent valves. I know from first hand experience that the Hyundai Elantra engines always bent valves when the timing belt failed and wouldn’t waste my time testing one.

As a diagnostic tool, A compression test is simpler to run than a leakdown test and gives good relative conditions cylinder to cylinder. A leakdown test will tell you just how bad the engine is overall and relative cylinder to cylinder. A consistent cylinder to cylinder reading shows nothing is bad with any specific cylinder.

A 20% leakdown on all cylinders tells you this engine is pretty well worn and putting out less HP that when new. It also tells you a valve job or ring job with no overbore would bring this engine right back to spec. Just another tool in the arsenal to diagnose the problem.

Some folks says,

I don’t see how a leak down test could be done properly with camshaft timing being out.

If the engine is still properly timed and the timing belt is still intact and in place then if he did the leak down test correctly you might find problem with cylinders/valves.

When the belt breaks, the camshaft will always have the valves open on one or two cylinders. This will show up on a leak-down test as valve damage, when the valves are simply being held open by the cam because they’re out of sync with the pistons.

Thanks for sharing.

Again, a leakdown test can be done with no timing belt on the engine at all but the cams will have to be rotated manually to test each cylinder. That means rotating the camshaft or shafts independent of the crankshaft.

Your engine is non-interference; meaning valve and piston damage will not occur with a broken or jumped belt.

Your engine will not suffer valve to valve damage because one cam is belt driven and the other cam is synced to it by gear.

When the belt breaks, the camshaft will always have the valves open on one or two cylinders. This will show up on a leak-down test as valve damage, when the valves are simply being held open by the cam because they're out of sync with the pistons.
No, that's not right.

With a broken belt, take off the valve cover. Loosen either the rocker arms, (OHV) or the camshaft mounts (OHC). Now, all valves are closed, regardless of timing.

You CAN do a leakdown; you CAN'T do a standard compression test.

(I think whoever's steering you away from the leakdown test simply doesn't have the equipment to do it, and doesn't want to admit it.)

A vacuum test requires the engine to be running. A compression test requires the engine and camshaft to be sync’ed together and the engine able to turn. But a compression test is done with the engine off. A leak-down test only requires the cylinder being tested to have the valves closed. With a broken timing belt or chain, the camshaft can be manually rotated to do this, or lifters and followers can be loosed to achieve this.

Wish you happy holidays. I find the following video about bent valve.

Suzuki Cylinder Leak Down Test - Part 2 (Are Valves Bent?)

This video talked about to have timing belt on to do Leak Down Test. He ask to keep Crankshaft aligned with timing mark and putting cam sprockets (Exaust, Intake)back in time . To make this happen, he put timing belt on.

As meanjoe75fan suggests, will the above condition happen without timing belt on.

Another mechanic suggest to do timing belt job first (on timing belt break event for non-interference engine), then go for Leak Down Test to confirm engine health.

Is it a right way to do it?

Thanks for your sharing.

With the timing belt off the cam and crank will turn independently. That means the cam can be positioned to be in time without the belt. But then both pulleys are free to move and get out of time and that is what causes problems. Air pressure will force a piston down to bottom dead center and that throws two other pistons up.

The mechanic in the video has patiently moved through all the cylinders but I wouldn’t go beyond one cylinder. The cylinder head will come of whether there is one bent valve or sixteen. But as I mentioned earlier, air pressure will move the piston down and depending on whether it turns clockwise or counter clockwise some or all the valves will be open if the timing belt is in place.

Thanks for your suggestions.

I have another question. Do we able to use OBD reader/auto scan to detect timing belt failure (without opening the timing cover)?

What is the code shows up in reader?

Thanks for sharing.


The answer is no

There is no scanner or code reader that conclusively tells you the timing belt has failed

The sound of an engine being cranked with no compression is very identifiable. And jumped time is by far the greatest cause for the simultaneous loss of compression on all cylinders.

A compression test is a dynamic pressure quick test to determine if there’s problem with a cylinder. But it doesn’t indicate what the problem is and sometimes doesn’t indicate a problem at all.

A leak-down test is a static pressure test of each cylinder and identifies where the problem is.

For example, I had a Jeep Wrangler come in with misfire codes for cylinders 3 & 4 with a 4.0 liter engine. After checking all the obvious things that can cause misfires, I did a compression test on those two cylinders and the readings were within spec. I then performed a leak-down test on those two cylinders. And when I applied compressed air into cylinder 3, air came out of the spark plug hole for cylinder 4. And when I applied air into cylinder 4 air came out of cylinder 3.

What had happened was a small crack developed in the head gasket between cylinders 3 & 4. And when I performed the compression test, the piston moved up and down fast enough to build enough pressure on the gauge to read normal. The crack was too small to effect the readings. And a compression gauge only shows max pressure from that cylinder. It doesn’t leak down to indicate leak.

But with the static pressure leak-down with constant air being applied to the cylinders, you can pretty quickly figure out where the problem lies.


A leak down test on your 1999 camry is perfectly good to determine the problems with the engine. If the timing belt broke, it bent valves.

Now as far as a leak down test being adequate for every modern vehicle is argueable. Though i have a perfect understanding of how an engine operates, in some of the engines i test that have five valves per cylinder and multiple camshafts with variable timing, they dont always like a leak down test. I cant explain exctly how or when all the valves are closed. Obviously they must be at many points to make the engine run. but no matter how much we know about engines or how cars operate we will never understand it all. and hopefully learn something knew everyday we work on them. those who already know everything should just become engineers.

Now on your 99 toyota it is not that complex. and we do all know everything about it. im sorry.