If there's a difference between what an oven is set to and what an oven thermometer shows, how do I know which one to trust? Reviews of oven thermometers suggest that not all oven thermometers give accurate readings. The reviews usually involve using three or four different thermometers and/or specially calibrated scientific precision instruments to gauge accuracy, but most home cooks / bakers don't have access to precision instruments or multiple thermometers. So I'm not sure how to determine whether my thermometer is accurate.

Here's the full story. I recently purchased a countertop Hamilton Beach toaster / convection oven. Dishes I prepare in it take significantly longer to bake/cook than the stated times. If something says "bake for 35 minutes or until a toothpick," etc., it usually takes 45 minutes or so for the dish to be done. I suspected the oven might be running cold, so I purchased a Rubbermaid oven thermometer to check. It's the first oven thermometer I've owned.

I tried the thermometer in the HB oven on both the convection and regular settings. I also tried it on the oven in my fifteen year old GE electric range. According to the thermometer, the GE oven is accurate: if I set it at 350, the thermometer shows 350. The Hamilton Beach one is apparently too hot by 10 degrees or so, on both regular and convection: if I set it at 350, the thermometer shows 360. I was surprised at this, because as I said the catalyst for my purchasing the thermometer was that I thought the Hamilton Beach might be running colder than the chosen temperature.

I've noticed that on the GE too I have to bake things for longer than the stated time, though to a lesser extent. I've always assumed it's just because recipes are conservative and err on the side of not burning food. If there's a range of times ("Put in a 350 degree oven for 45 to 50 minutes"), cook times on the GE are usually at midpoint to higher end of that range.

The Hamilton Beach does give off heat rather noticeably when in use. I'm wondering the higher actual temperature is an imperfect way for the manufacturer to compensate for heat escaping from the unit. As a countertop device, it's not anywhere as efficient as the very solid GE.

But I'm also wondering about the oven thermometer itself. After all, there's no guarantee that the thermometer isn't off! How do I know whether the oven thermometer is accurate? It's at least possible that the Hamilton Beach is entirely accurate, the GE runs ten degrees cold, and the thermometer readings are ten degrees too warm.

  • 1
    It probably wouldn't work with an oven thermometer, but I've tested kitchen thermometers by putting them in boiling water. Water boils at a constant temperature (100C at sea level) so it's a simple and accurate way to test a thermometer.
    – Ross Ridge
    Dec 10, 2016 at 1:23
  • Of possible interest: cooking.stackexchange.com/questions/37470/…
    – Jolenealaska
    Dec 10, 2016 at 2:14
  • 2
    The larger oven will store more heat (especially as being older the metal may well be thicker). So it will cool down less when you put the food in. This has a big effect on cooking times. It may also have a more powerful element. An oven thermometer is also going to be a bit less predictable if much of the heat is radiant.
    – Chris H
    Dec 10, 2016 at 9:23

2 Answers 2


Use a pot of water!

  1. Find out the lowest temperature of your oven (needs to be at least below 212 F for it to work). Preheat the oven to this heat
  2. Heat water on the stove to a temperature let's say, about 10 F below the oven temp. Use a candy thermometer and leave it in the water, but not touching the bottom of the pan.
  3. Once the oven is preheated, transfer the pot of water to the oven, and leave the candy thermometer on the whole time.
  4. Periodically check the temperature of the water. It should asymptotically approach the temperature of the oven. If the water temp closes in on the temp of the oven, the oven thermometer is calibrated!
  5. You can verify by checking the temperature on the opposite end. Warm the water on the stove to 10 F above the oven temp, and watch the candy thermometer go down to the oven temp.

Note: it is very important to get the water on the stove close to the oven temp. Water has a very large specific heat capacity and will take quite a while to heat up in the oven, as it takes a large amount of energy to heat water.

  • 2
    Why not put the candy thermometer besides the oven thermometer? While the heat capacity of water is a nice for buffering the temperature in the oven it also means that temperature equilibrium takes much time (even if you are presumably close by preheating) and the heat loss through evaporation can easily mess up this experiment (=> at least use a lid on the pot with water). Dec 20, 2016 at 15:52

Depending where you live, you may be able to get official proof of accuracy for as low as $25 ~ $40 equivalent. If you bought an expensive thermometer and want to be really sure, this might be a decent option. If you are in food business and, for example, need to know for sure if meat reaches the minimum required temperature for minimum required time, it's a must.

Sometimes you can ask them nicely and if they are not busy, or have students on training, you can get it tested for free - without a paper, then, of course, so that's only for private use.

Consider looking for measurement agency in your area. Most countries that have sanitary requirements about temperature also allows to somehow obtain papers on your measurement tools.

  • Thanks for the suggestion! The thermometer I purchased is cheap ($7 or so). Also, I don't understand Polish, sadly, so I could not tell what the link you provided was saying. I live in the US, and I have no idea how I would go about contacting an agency that provides official proof of accuracy. I'm not sure I can take this route.
    – verbose
    Dec 22, 2016 at 6:01
  • @verbose Link is just an example of agency that certifies thermometers for the price I gave - in my country. In other countries, well, comprehensive list would be oo long for Q&A. For $7 it has a little sense. For someone that really has to know for sure (like restaurants), it's another matter.
    – Mołot
    Dec 22, 2016 at 6:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.