Every one of these I've owned or seen has to be calibrated depending on the type of surface you're measuring. And then its only as accurate as you've calibrated it. I generally don't bother with calibrating them since I usually use them to do comparison readings - like tire temps. Example, you're checking four tires. It doesn't matter if it reads 50 deg or 200 deg - what I'm interested in is the "difference" between the four for comparison. But if you want to measure one thing and you want it accurate well, good luck. Generally I've found the variance between types of materials to be around 10 deg., which for most purposes doesn't matter. But if you're checking body temps for fever, a 10 deg. inaccuracy won't cut it !!