I've been thinking of getting a temp reader to monitor engine temps so I went to the local electronics store to test some out.

What I found was that even among the more expensive temp guns (RayTek MT4, QuickTemp by Linear Labs) which are $70-120, the temp reading against a fixed object varied by more than 10 degrees.

Question is...how do we figure out the "true" accuracy between these temp readers? I read somewhere about pointing it at boiling water, but was wondering if that was an accurate way to gauge the reading or are there other ways to do that?