Kawasaki Triples Club

from: http://www.3cyl.com/mraxl/manuals/maint ... htm#tuning

To check the compression, warm the engine to normal temperature, then remove the spark plugs. Use a gauge which registers 200 psi and insert the tip into a spark plug hole. Twist the throttle to the wide-open position, then crank the engine with the kickstarter until the gauge levels off at a maximum reading. NOTE: The compression process begins after the piston crown closes off the exhaust port during the upstroke. Repeat the process on all cylinders.
A good reading, on all H models is 142 psi. The S1/KH250's should develop 170 psi; S2's, 156 psi; the S3/KH400's, 155 psi. (Note: altitude correction factor may be required).

It probably depends on why you are compression testing but according to Dan (http://www.dansmc.com/compression_test.htm )

Technically, You should do this when the engine is hot, but I don't like burning my fingers, so I always do it when the engine is cold. A hot engine would probably read a few pounds more, but a cold reading will tell you what you need to know.

When I've checked my compression (KH400) I've done both hot and cold and there hasn't been a great deal of difference (about 5psi).

If you think about it for a wee while , it would make sense that compression does not vary much , hot and cold.

given that the internal combustion engine. Relies on a compressed spark , and that it needs to start, when cold or warm ( ! ). It has to stay constant . Engine designers choose the piston, rings and cylinder materials.and the engine clearance tolerances so that a constant tolerance is maintained throughout the working range of the engine.

If they didn't , you'd be heating an engine up to operating temperature. To get it started. Or stopping every moment to let it cool down as it exceeded a "cold start " running tolerance.

I warm engine with all the moving parts expanded , on a new engine with correct and "perfect" tolerances , will seal better , hence a minor improvement in compression. But , with the poor accuracy of single digit variances. On the Testers we tend to use ; much of any improvement in the reading could , and probably is influenced as much by the warmer air passing thought the Compression Tester ?

and as I have been typing this, I. Am thinking , " this is so bleeding obvious , I feel stoopid typing it !! "

but hey oh, cue a torrent of contradictory theory no doubt ??!!

For accurate comparison, the engine temp just needs to be consistent. especially when warm. I check with the engine cold personally.