Thermometers: How They Work, How To Select Them, and How To Use Them

Nobody knows how many millions of dollars are wasted on overcooked food, but far more importantly, the US Center for Disease Control (CDC) estimates that in 2011 roughly one in six Americans got sick from foodborne illnesses, about 128,000 were hospitalized, and 3,000 died, about the same number who died in the attacks in 2001 or Pearl Harbor in 1941. The difference: Many were children.

Digest this: The fever, sweats, and runs that people call “stomach flu” are almost always a food borne illness caused by bacteria. True stomach flu is a viral infection caused by a virus and is exceedingly rare. Chances are that if you think you had stomach flu, you really had a foodborne illness that probably could have been prevented by proper cooking. Read this article by health writer Serena Gordon of HealthDay about her brush with death.

Thermometers are as important as knives and forks. Only knuckle draggers think thermometers are for sissies. I want my food safe, tender, juicy, and flavorful. The temperature of the cooker and internal temperature of the food controls all of these things. Understanding optimum and safe temperatures is at the core of good cooking. Proof: Of the thousands of barbecue teams competing for prize money every weekend across the nation, I have never ever seen one that didn’t use a digital meat thermometer (usually a Thermapen), and the majority also use digital oven thermometers. USDA, Consumer Reports, and this website are unanimous: If you use a good digital thermometer and handle food properly, you can reduce the risk to zero.

Below I explain how thermometers work, why some are better than others, why you cannot trust the old ways of cutting into the meat, and I debunk some myths. If you want to cut to the chase and just buy a thermometer, click here to go to our recommendations and to search our database of test results, ratings, and reviews. Our database contains accuracy tests, ratings, and reviews of more than 200 thermometers. We use sophisticated NIST certified lab testing equipment. We even highlight our favorites and best buys. There is nothing like it anywhere.

How we test thermometers

Our thermometer tester, Bill McGrath is an electrical engineer, and he uses specialized equipment that measures how accurate a thermometer is and how long it takes to get to an accurate reading.

Here’s a video of how we test thermometers:

What you need to know about thermometers

Alton Brown, Food Network Star and author of multiple cookbooks says “Bimetal coil thermometers are about as accurate as a sniper scope on a nerf gun.” Don’t believe him? Look closely at the photo at the top of this page. Some bimetal dials are off by about more than 100F!!!!!

Most grills and smokers come with bi-metal dial thermometers, and they’re usually crap. It is not unusual for this design from the 1800s to be off as much as 50°F like the one above (on an expensive and otherwise superb grill). You cannot trust them. I have readers tell me that when they bought a good digital from my list below that they learned their grills were off by as much as 100°F! This is a recipe for well done steaks, late meals, cold food, embarrassment, shame, and ostracism.

Worse, handheld “instant read” bi-metal dial food thermometers can take up to 30 seconds to read accurately. Digitals can read in 1 to 6 seconds with much greater precision! Don’t take my word for it.

Cooking without good digital thermometers is like driving at night without headlights. Spend the money for good thermometers or you will spend the money on ruined food later! They will pay for themselves by saving your meat and your face.

Without good digital thermometers there’s a good chance you’ll be making lame excuses for overcooked meat, undercooked meat, or, worst of all, apologies at bedside in the hospital as your guests recover from food-borne illness.

Here’s an overview of the types available (more on them below):

You can’t tell by poking your hand

“The idea that you would rely on intuition to judge something you are terrible at judging makes very little sense to me. Why don’t you blindfold yourself too?” Nathan Myhrvold, food scientist, author of Modernist Cuisine: The Art and Science of Cooking and Modernist Cuisine at Home

A lot of cookbooks tell you that you can tell when meat is ready by poking it and comparing its resistance to the flesh on your hand. This is utter nonsense!

The resistance of the steak is going to depend on what cut of meat you are poking (sirloin is stiffer than filet), the grade of meat (prime is more tender than select), how thick it is (thick cuts will yield more than thin), the age of the steer (young is more tender), the breed of steer (cooked Wagyu is more tender than Holstein), the age of the meat (wet aged is more tender than fresh killed), and what the steer was fed (corn fed is usually more tender than grass fed), among other things.

In addition, the resilience of our hands differs from young to old, from thin to fat, from exerciser to couch potato. Why do so many cookbook authors repeat this bunk?

Yes, steakhouse chefs can tell a steak’s internal temp just by poking it. But they have poked thousands of steaks, all from the same supplier, all the same thickness, all cooked at the same temp. And I have news for you. Look carefully at the pockets of a top chef. There’s usually a thermometer peaking out.

For home cooks, there simply is no substitute for a good digital instant thermometer like the ones I recommend in our Buying Guide to Thermometers.

You can’t tell by holding your hand over the grill

Here’s a direct quote from a website I otherwise respect: “To check the temperature, hold your hand four or five inches away from the grill. You should be able to keep your hand there for four to five seconds. Less than that and it’s too hot; longer and it’s not hot enough. Note that certain veggies prefer a medium grill, which corresponds with 350 F or six to seven seconds of holding your hand over the grill.”

Really? You think we all respond to heat the same way? Then why is my wife constantly piling on more blankets while I shed them? Why can some people snuff out cigarettes in their hands?

I don’t care what the TV chef said, you absolutely positively without doubt no way no how cannot tell anything about the temp of a grill by holding your hand over the grate and counting “1001, 1002, 1003″ until until your palm starts to smoke. Each of us reacts differently to heat, and the heat 1″ above the grate can be significantly different than 6” above. You need a thermometer!

You can’t tell by the color of the juices

Scores of cookbooks tell you that chicken or turkey is ready when the juices are clear. That is simply not true. It may have been true once upon a time, but modern production methods have made this old wives’ tale false.

Modern chickens are grown so rapidly that the ends of the bones don’t calcify thoroughly, and so blood from the marrow, and that’s where blood is made, can seep out and tint the nearby meat even though the meat is cooked well past safe temp.

As for the juices, scientists tell us that myoglobin, a pink protein liquid, can tint the juices depending on the acidity of the meat. I have written a whole article explaining the issue and debunking the “chicken is done when the juices run clear” myth.

You can’t tell by cutting the meat

A lot of weekend warriors cut into their meat to check the color for doneness. You cannot tell by looking at the color of meat.

The problem is that the color you see on the grill is not the color that you will see on the table. That’s because the color in the cut changes as the meat absorbs oxygen.

All meats, including fish, have myoglobin in the muscle cells. Myoglobin is pink in most animals, and it remains pink and runny after contact with air. When heated it turns tan and thickens, and that is why medium rare meat is reddish pink, and well done meat is tan. But myoglobin also contains an iron compound called a heme and when it comes into contact with air it changes color in the same way that iron rusts in the presence of air. So when you make a cut into a steak it may look perfectly done to you, but as the myoglobin absorbs more oxygen it can turn brighter red.

In the photo above we see two slices of Iberico pork from Spain cooked to a safe temp (yes it is pork). The bottom one was exposed to air for about 10 minutes after I carved off some slices for dinner. As it was exposed to oxygen, it turned brighter red making it look medium rare. The top one was sliced from just behind it, moments before the picture. You can see it is less red, more pink, looking as if it is medium, which is what it was cooked to.

Worse still, the color of food is altered by the light you are using. Incandescent light is yellowish orange, fluorescent is greenish blue, most LEDs are slightly blue. Women who wear makeup know they always look better under incandescent light than fluorescent. In other words, the type of bulb you are using impacts the color.

Here’s what the US Department of Agriculture (USDA) says: “The color of cooked meat and poultry is not always a sure sign of its degree of doneness. Only by using a food thermometer can one accurately determine that a meat has reached a safe temperature. Turkey, fresh pork, ground beef or veal can remain pink even after cooking to temperatures of 160°F and higher. The meat of smoked turkey is always pink.”

Bottom line, if you really want to know when the meat is done to your likeness, you need a good digital thermometer.

A thermometer will not make your meat dry

Meat is about 75% water. It is not a balloon. When you stick in a probe a few drops of juice may escape but it doesn’t go pfffffft like a baloon and deflate. In an 8 ounce steak 6 ounces are water. If you lose 1/4 ounce, and you probably won’t, there are still 5 3/4 ounces of juice left. That’s 95%. Besides, much of the juiciness feel of a piece of meat is the melted fat and collagen. So stop worrying.

Where to stick it

Most grills come with cheapo thermometers that are an afterthought to manufacturers and are usually bought bulk from cheap producers. To make matters worse, the probe is located high in the dome well above the food. Unless you plan to eat the dome, this is a bad place for the thermometer. You need to know the temperature where the food is. So put the probe about 2″ to the side of the food (it is cold and there is a cold air bubble around it), and about 1″ above the grate (it is hot).

Some thermometers, like the Maverick, come with a handy clip that does the job just fine.

If you don’t have a clip, use a ball of foil. Make sure the tip, where the sensitive parts are, protrudes from the foil.

Listen to this email from a reader “I had been clipping the probe on the underside of the upper rack. That had the probe about 3″ to 4″ above the top of the meat. My food was taking much longer to cook than your recipes say. So I tried the probe in that location for about 1/2 hour and then moved the probe to the cooking surface, clipping it to the cooking grate about 4″ to 5″ below the previous location. The difference in temperature was about 25 to 30°F cooler at the cooking grate location! I never would have believed it! So in actuality, when I thought I was cooking at 225 to 230°F, I was actually cooking at 195 to 200°F! No wonder everything was undercooked!”

Just how hot is it?

When measuring meat temp, take the temp in more than one location because the composition of the meat and the unevenness of the grill can fool you. Insert the tip of the thermometer into the thickest part of the meat and go past the center. Slowly pull it out. Read the coldest temp. Test other locations. Here are two other tricks for taking the temp of a thick cut of meat.

1) Insert the probe from the side as in the porterhouse here. Stay away from the bone which heats at a different rate than the muscle.

2) Another option is to line up the probe tip on the outside of the meat (that’s a brisket below) until the point is past the middle. Then slide your finger tip until it touches the top of the meat (top photo). Now slide the probe into the meat until your finger touches the meat. The tip will be past the center (bottom photo). Now pull it back slowly and read the lowest temp. If any modeling agencies need a hand model my wife is available.

Different types of digital thermometers

Temperature Measurement Comparison Chart From National Instruments

Thermocouples: The best, especially for rapid read food thermometers. Thermocouples are the best rapid read food thermometers because they’re fast and precise, with a small sensor. Their margin of error can be less than 1°F. Thermocouple probes have two tiny wires of different metals welded at the tip, most often nickel and chromium (called Type K – described elsewhere on this page). The heat causes a tiny voltage to appear across the dissimilar metals which are connected to a meter that measures the voltage and calculates the temperature.

Thermistors: Good for continuous readings for large roasts and oven thermometers. Thermistors are usually not as quick as thermocouples, they tend to be thicker, and they can be slightly less accurate, usually with a margin of error of 5°F. They are best for leaving in large roasts and oven for continuous readings. Thermistor units send a current through a wire in the probe with a resistor in the tip. Its resistance to the electrical flow changes with temperature and the meter measures the voltage across the resistor.

Resistance Temperature Detectors (RTDs): For leaving in roasts and ovens. RTDs are are the most stable and accurate but they have a slow response time.

IR (infrared) sensors. Infrared-sensing thermometers usually look like guns and they measure the energy radiated from a warm surface and convert it to a temperature reading. Some surfaces emit more infrared at a given temperature than others, so some units allow the user to adjust the emissivity setting to fine-tune the accuracy of the thermometer. This should rarely be necessary unless very high accuracy is needed. These units are useful for determining the temperature of a cooking surface like a skillet or griddle.

Other features, specs, and technical terms

Adjustable. The accuracy of some thermometers can drift, and these thermometers can be adjusted to bring them back to the correct reading.

Alarms. Some devices can be set to alert the cook when a high or low temp is reached either with an audible alarm such as a beep or a visible alarm such as a flashing display.

App. Bluetooth and Wi-Fi enabled devices have a smartphone or tablet app that talk to the device.

Auto shutoff. If you don’t use the device for a set period of time, it will shut itself off to save batteries.

Backlight. Digital thermometers are hard to read at night unless they have a backlight.

C/F Switch. Most thermometers have a switch to change the display from Celsius to Fahrenheit.

Liquid filled thermometers. Old-fashioned liquid filled thermometers are very small glass tubes filled with a liquid that sits in a bulb at the bottom. As it warms, the liquid expands. These thermometers are slow but they can be very accurate. Because they do not need batteries, they make good refrigerator and freezer thermometers, but they cannot read a small area such as the center of a hunk of meat well.

Logging. Some thermometers can remember the readings taken over time and create a log that can either be printed or exported to a spreadsheet.

Margin of Error. Most thermometers are considered to be accurate if they are within plus or minus 3°F of the target temperature. High end thermometers are more precise than this.

Min/Max. The minimum and maximum temperatures it is capable of reading. On some devices, especially those which use different probes, the Min/Max can vary with the probe.

Speed rating. We measure how long it takes the thermometer to go from 32°F to 211°F and from 212°F to 33°F. But you have to be careful about the time manufacturers quote. Often they use an industry standard called “time constant.” That is the time it takes to get to 63% of a full reading, and a full reading takes five times that. So if they say the time constant is 0.6 seconds, as does the manufacturer of the Thermapen, the unit will be precise with a full reading in about 3 seconds.

Another factor to consider is how fast the display refreshes itself. The Thermapen refreshes every 0.5 seconds. This means you can slowly insert it and remove it and it will give you a new reading every 0.5 seconds. The Thermoworks K-type Fast Response Meat Probe #113-151, which can be plugged into different meters, is slightly faster with a time constant of 0.5 seconds and precise read in 2.5 seconds. But if you plug it into the MTC meter, which refreshes every 1 second, the probe is actually faster than the meter, and combined they are slower than the Thermapen. Another factor is the conductivity of the medium you are measuring. Food, which consists mostly of water, reads faster than bread, which consists mostly of air. That is because water is a better conductor than air.

Water resistance rating. Some manufacturers simply specify that a thermometer is “water resistant,” but this is not a regulated description and should be taken with a grain of salt. Some manufacturers use a third party service, the International Electrotechnical Commission (IEC). It rates devices using an International Protection rating code (an IP code). An IP code might look like this: IP65. The first digit quantifies protection against solids like dust, and the second digit quantifies protection against liquids. The digits range from IP00 to IP68. If a thermometer is rated IP65, it was tested and found to be completely protected against dust as well as protected against low pressure jets of liquid from all sides. So it is OK in the rain, but NOT protected against a swim in the BBQ sauce.

Other thermometer types

Liquid filled thermometers: Good for refrigerators and freezers

Old-fashioned liquid filled thermometers are very small glass tubes filled with a liquid that sit in a bulb at the bottom. As the liquid warms it expands. They are slow but they can be very accurate. Because they do not need batteries, they make good refrigerator and freezer thermometers, but they cannot read a small area such as the center of a hunk of meat well.

Popup thermometers: Unreliable

Popups have a compound in the tip that melts at a determined temp and releases a spring that pops the stem up. Although they can be accurate, they can also stick, they read only one part of the turkey, and they are usually set too high to prevent litigation. Pop-ups are why your turkey tastes like cardboard. Throw them out.

Bi-metal dial thermometers: Most are not reliable

Most bi-metal coil dial thermometers mounted in grill hoods should be called heat indicators, not thermometers. We do not recommend them. They have round clock-like readouts and the sensor uses two strips of metal bonded together and rolled into a coil. Each metal expands at a different rate, turns a shaft, and this provides the reading on a dial. Bi-metal meat thermometers can take up to 30 seconds to give a final reading and because the sensor can be 1/2″ long or more they cannot read a specific location in meat. Most thermometers built into grills and smokers are bi-metal, but they are often low quality in order to keep the grill price down. They can easily become unreadable if they fill with smoke and or water. Also, these grill thermometers are mounted in the dome where the temp can be very different from the temp at the cooking surface making them both unreliable and misleading.

If you get condensation or water under the glass of a bi-metal thermometer, put it in a zipper bag with a couple of cups of rice or dried pasta and seal it up. In about a week the grain will have absorbed the moisture and your thermometer should be back to normal.

Now that I have slammed bi-metals, I should tell you that there is one brand that a lot of the pros use on their big pits and they tell me it is pretty accurate: Tel-Tru. It comes with a wing nut, a range from 150°F to 700°F, 1.75″ dial, and a 2.13″ stem. As good as they are, that dial face is just not as accurate as a numeric readout that is precise to within a degree or two.

Thermometer calibration

Buy good therms and take care of them. Here are some tips on caring for them.

Calibration. You should check a thermometer’s accuracy soon after you buy it, then once every year, and again if you drop it. You can check your thermometer’s accuracy with boiling water and with ice water.

Boiling water. Bring a pot with about 3″ of water to a boil and insert the probe. It should read about 212F. Notice the key word “about”. The exact reading can vary slightly with air pressure (factory calibration is based on one atmosphere, about 30″ of mercury). Minerals in tap water can cause minor variations, so use distilled water if you want to be absolutely precise. I just use tap water. Remember that water boils at lower temps at high altitudes. The ThermoWorks website has a nifty calculator that helps you determine what your boiling point is.

Ice water. Fill a tall glass with ice cubes, not crushed ice, add cold water, and let it sit a minute. Insert the probe and make sure the tip is not below the ice or touching the ice. The temp below the ice can be several degrees above 32°F (0°C) and the temp of the ice can be below 32°F. The experts at ThermoWorks say “Make sure the probe is in the middle of the ice water mixture and then gently stir for best results.” The ice water test does not vary with altitude.

Common malfunctions, troubleshooting, and some important warnings

Lately I have been noticing comments on Amazon from disgruntled thermometer buyers complaining their units have failed. They are mostly about thermometers with flexible cables. As I read their comments carefully, I have come to the conclusion that most of the problems are because the owners did not read the instructions.

Failures are usually probe or cable failures rather than electronics failures. With proper care, probes and cables should last years. I have numerous braided cable thermometers (I test a lot of thermometers) and only one has failed and then it was after several years of abuse. I was able to buy a replacement at a reasonable price.

I raised the subject with a bigshot at Maverick and he told me this story: “I had a consumer drive to our office. He advised he got HHH on his probe [an error message]. I assembled the unit and turned it on. I saw the HHH. I pushed the probe jacks into the meter a little harder to make sure they were making good connection. Solved problem. Consumer stood in my office with mouth open.”

Apparently this is a common problem. Make sure you have crammed the jack all the way into the meter and twist them back and forth so they make better contact.

If the unit still displays an error or cryptic message like HHH or LLL, let the probe tip come to room temperature. The error may have been because you exceeded the range of the probe. If the problem still persists, remove the batteries and put them back in so the device can reset itself. If there is still a problem, it is likely the probe wire has shorted out, not a failure of the electronics.

Use the right probe. K-type probes, which are usually yellow plastic and have two flat spade connectors, are interchangeable and work on any thermocouple meter that takes K-type probes. Alas, this is not true of most other probes. Although many thermistors use minijacks like the ones for earbuds, they cannot be interchanged from one meter to another.

Keep the probes dry. Thermometer probes with braided cables like the ones on the otherwise excellent Maverick ET-732 or ET-733 occasionally fail if water gets into the place where the solid probe and the braided cable connect. That means you absolutely cannot submerge them when you wash them. The braid can get wet in a drizzle because the wires within are coated, but the junction between the braid and the solid probe is vulnerable.

Stay under the temp rating. Therms with cables tell you in the manual what the max temp is. Pay attention. According to the AmazingRibs.com science advisor Prof. Greg Blonder, “Almost all defective cables are the result of exceeding the upper temp limit. Most consumer cables are sheathed in teflon, which melts at 500°F. Even if the air in your grill is only 350°F, just quickly touching the grill surface can melt the teflon because it can be 600°F. I often loosely wrap the cable in foil to provide air insulation, and when I’m feeling fancy, slip a braided stainless tube over the teflon.”

Keep the tips clean. Carbon can build up on the tip of an oven/grill/smoker probe and insulate the sensor giving a false reading. To clean a smoker probe, use a soapy sponge and focus only on the 1″ near the tip since the sensor is usually within the lower 1/4″. You don’t have to clean the smoker probe after every use, just don’t let it build up an opaque coat.

After using a meat probe, get a sponge wet with soapy water and sponge off only the tip that was in the meat. Then rinse under the faucet or with a wet paper towel being careful to keep water off the braid and especially out of the junction.

Order spares. Probes are like hard drives. They can fail and they will fail at the most inopportune times. Fortunately most are cheap. If you depend on your thermometers, and you should, keep a spare probe or two in inventory because you know for sure yours is going to die on Christmas Day when you are cooking a 16 pound Wagyu Prime Rib that cost you $400.

Don’t drop them. DOH!

Don’t smash or crimp the cable. If the lid on a grill smashes the cable it can break the internal wire. Crimping it can also break the internal wire.

Don’t leave them out in the rain. Many of the best digitals are not water resistant, no less water proof. If it looks like it may rain, put your meter in a zipper plastic bag.

Stay within the temp ranges. Pay attention to the specifications in the manual. Each unit has a temp range. Don’t go over it! You can damage the probe.

All this raises the question: Why can’t manufacturers come up with a fully submersible probe and cable? And those manuals! Would you please hire a professional writer rather than asking your English major daughter to write it? C’mon guys! Get with it!

Problems with induction cooktops. New induction cooktops work with electromagnetic fields and they can interfere with some thermometers that are not well shielded. The workaround is to momentarily turn off the cooktop when you take the temperature of something above it. This short interruption in the cooking process will not harm the food or significantly slow the outcome.

Order one of our award winning Temperature Guide Magnets

We now have two versions of our award-winning temperature guides. These unique award winning guides show you both USDA recommended temperatures for all your favorite meats as well as the temperatures recommended by chefs (they are not always the same). The larger one contains more foods and other temperature benchmarks such as when fats melt, when collagen melts, oil smokepoints, sugar stages, and more.

Comprehensive Food Temperature Guide (8.5″ x 11″, above) which sells for $9.95 (it is free if you join the Pitmaster Club).

Click here to visit our comprehensive database of reviews and ratings of thermometers.

Grommets (but no Wallaces)

One of the problems with a lot of grills is that you have to thread thermometer cables under lids, down chimneys, through vents, and then find a way to attach the probe to the grates without letting the tip touch the metal. The lids often crimp or cut the cables and you’re never quite sure where the tip of the probe is and if it is touching meat or metal. Then you forget about it, lift the lid, and you thermometer goes flying into the neighbor’s yard.

The solution is to drill a hole just above or below the cooking grate and insert the probe through the hole. Make the hole just a bit larger than the probe, and don’t worry if it leaks a little. That small amount of leakage won’t hurt anything. If you wish, look for silicone grommets at your hardware stor, from Grainger or from McMaster. Not all probes are the same diameter, but 1/8″ should work for most.

Our science advisor, Prof. Greg Blonder, says “The steel is very hard. Create a dimple to help center the drill by hammering a concrete nail or punch into the side. I also cut a small block of wood to fit behind the spot you are drilling, clamped or held tight. Helps prevent shredding on the inside of the drilled hole, and drops vibration while drilling. Use a newer, sharp, high speed bit.”