How to Use an Instant Read Thermometer and Which One to Buy

An instant-read thermometer is a small digital probe you push into food for a few seconds to find out how hot the middle is. It is the one tool that turns a guess about doneness into a number. This page is for the cook who makes dinner most nights and wants to stop cutting into chicken thighs to check them, and for the cook trying to decide whether a $20 model or a $100 one belongs in the drawer.

Most people get one thing wrong, and it isn't the brand. They assume the thermometer does the measuring. It doesn't. The thermometer reports the temperature of wherever its tip happens to be, and putting the tip in the right place is your job. A fast, expensive probe sitting against a bone gives you a precise answer to the wrong question. A cheap probe placed well gives you dinner.

There are three kinds of food thermometer in home kitchens, and the USDA's Food Safety and Inspection Service lays them out plainly. The old dial type uses a metal coil that senses heat along a stretch of its stem, so it has to go 2 to 2.5 inches into the food before it tells you anything useful. Digital thermistors sense heat with a small resistor bonded into the tip, which is why FSIS says they can read a burger patty as well as a roast. Digital thermocouples, which FSIS calls the fastest and most versatile of the three, read in about 2 to 5 seconds.

That difference in where the thermometer senses heat matters more than any spec on the box. A dial thermometer can't read a pork chop that is an inch thick, because its sensing stretch is longer than the chop. A digital probe with its sensor at the tip can, as long as you put the tip in the center.

On accuracy, FSIS says most consumer food thermometers land within about 2 to 4°F of the true temperature. Penn State Extension sets a tighter bar, saying thermometers for food use should be accurate to plus or minus 2°F. The premium models do better than either: SmokingMeatGeeks measured the Thermapen ONE at ±0.5°F in controlled water-bath testing.

For weeknight cooking the choice is really between the two digital types, and nearly every folding or pocket model sold as an instant-read is one of them. The dial thermometer has a place in a pot of soup. It has none in a skillet.

Here is why you bother. FSIS states that the color of meat and its juices is not a reliable sign of doneness or safety, and it cites a 1995 Kansas State University study in which a significant number of ground beef patties turned brown before they reached 160°F. Cooks who judge by color get caught both ways: meat that looks done and isn't, and meat that is cooked well past done because it didn't look ready. Time fails for the same reason. A supermarket chicken thigh varies in thickness from pack to pack, and the recipe's 25 minutes doesn't know which pack you bought.

1. Check the thermometer in ice water once, before you trust it. Fill a glass to the top with ice, crushed if you have it, then add water slowly until it sits about half an inch below the top of the ice. ThermoWorks says to let it stand for one to two minutes and stir while you test, so the tip isn't resting against a single cube. Hold the tip in the middle of the glass, away from the sides and bottom. FSIS says a working thermometer reads between 31.1°F and 32.9°F. Cost: about three minutes and a tray of ice.

2. Look up the number before the food goes in the pan. 145°F for a steak or pork chop, 160°F for a burger, 165°F for any chicken, per the FSIS safe temperature chart. Knowing the target means you check once or twice near the end instead of stabbing the meat every few minutes.

3. Find the thickest part and aim the tip at its center. FSIS tells you to go into the thickest part and avoid bone, fat and gristle. On a bone-in chicken thigh, that means the meaty middle, not down where the bone runs. You'll know you placed it right because the number settles instead of jumping around. Cost: nothing.

4. On anything thin, go in from the side. A digital probe senses at its tip, so on a chop or a boneless breast, push it in horizontally through the edge until the tip sits in the middle of the meat. Going straight down from the top puts the tip either too shallow or through the other side and onto the pan. This is the whole reason thin chops dry out: there is so little margin between underdone and overdone that you need the tip exactly in the middle to see where you are. The same horizontal entry works when searing a strip steak in cast iron.

5. Wait until the number stops climbing. An instant-read isn't instant for every model. FSIS puts thermistors at about 10 seconds to a stable reading and thermocouples at 2 to 5. Watch the display: when the last digit stops moving up, that's your reading. Cost: somewhere between 1 and 10 seconds depending on what you own.

6. Check a second spot. Penn State Extension says to check at least two locations. On a whole bird, FSIS names three: the innermost part of the thigh, the innermost part of the wing, and the thickest part of the breast. On a tray of thighs, check the biggest one, and one more.

7. Wipe the probe before every new check. A probe that went into raw-centered chicken shouldn't go straight into the finished piece next to it. A paper towel does it.

8. Let whole cuts rest. FSIS requires steaks, chops and roasts of beef, pork, lamb and veal to rest at least 3 minutes after they reach 145°F, and it treats that rest as part of the safety step, not just a way to keep the juices in. If you're working through thick supermarket chops, the USDA 145°F rest is where the timing comes together.

The reading is far higher than it should be. The tip has almost certainly gone through the food and is touching the pan, or it's resting against a bone. Penn State warns against letting the probe touch the sides or bottom of a pot or pan. Pull back, move the tip toward the center and read again.

The reading is lower than it should be, and it keeps changing. You read it before it settled. On a thermistor model that can take about 10 seconds, per FSIS, which feels long when the skillet is spitting at you. Hold it there anyway.

You took the temperature after the food came off the heat. FSIS is specific that the 145°F reading for whole cuts is taken before removal from the heat source. Measure in the pan, then move it.

The meat looked done, so you skipped the check. This is the mistake the 1995 Kansas State study was about. Browned ground beef can still be under 160°F, and FSIS calls a thermometer the only reliable way to know harmful bacteria have been destroyed.

You're using a dial thermometer on a burger. A dial needs 2 to 2.5 inches of stem in the food, per FSIS. No patty is that thick. The reading you get is partly the air above it.

The thermometer reads 36°F in ice water. FSIS's acceptable range is 31.1°F to 32.9°F. A dial thermometer can be adjusted with the nut under its head. Most digital instant-reads can't, according to both Ask USDA and Penn State, which means a digital that fails the ice test needs replacing, or at least needs you to mentally correct every reading from then on.

Water got inside. Not every thermometer handles a rinse well. SmokingMeatGeeks reported moisture intrusion in a Javelin Pro Duo during routine cleaning checks. If your model isn't sold as waterproof, wipe the probe rather than running the whole thing under the tap.

Ground meat versus whole cuts. A pork chop is safe at 145°F with a 3-minute rest. Ground pork needs 160°F. FSIS explains the gap: grinding spreads surface bacteria through the whole batch, so the middle of a burger isn't protected the way the middle of a steak is.

Poultry, in any form. FSIS sets 165°F for all of it: whole, parts, ground, and stuffing cooked inside the bird. A bone-in thigh for a weeknight braise has the same target as a boneless breast. Thighs are forgiving at that number; breasts are not, which is why the second check matters more on white meat.

Fish and leftovers. Fish is 145°F. Leftovers and casseroles are 165°F all the way through, and a thermometer is the only way to know the center of a reheated dish has caught up with the bubbling edges.

Altitude, if you calibrate with boiling water. Water boils at 212°F at sea level. Ask USDA notes that high altitude lowers the boiling point by roughly 2 to 5°F or more, so a boiling-water check in the mountains will look like a faulty thermometer when it isn't. Use the ice-water method instead; ice melts at 32°F wherever you live.

Hot days and food left out. The thermometer is useful after cooking too. FSIS says hot food should be held at or above 140°F and cold food at or below 40°F, and that perishable food shouldn't sit out more than 2 hours, or more than 1 hour when it's above 90°F outside. At a cookout, a quick check of the potato salad settles the argument.

The Thermapen model year. ThermoWorks announced the Thermapen NANO as the successor to the Thermapen ONE on August 20, 2026, and Homebrew Finds reported that the ONE is being discontinued. Anything written before this autumn recommending the ONE was right at the time; whether you can still buy one depends on closeout stock.

My position is simple: most weeknight cooks need a good cheap thermometer, not a fast expensive one, because the difference between them is a couple of seconds and the thing that ruins dinner is placement.

The best case for spending more is a fair one. When you're checking a chicken in a hot oven with the door open, every second counts, and a probe that reads in one second means less heat lost and fewer burned knuckles. Food Network's review of the Thermapen ONE found it gave a full reading in one second every time, read 165°F on chicken, 32°F in ice water and 212°F in boiling water, and is waterproof for 30 seconds at 39 inches. That's a real instrument. But at a skillet on the stove, the gap between one second and three is the time it takes to read the display. SmokingMeatGeeks' Nate Crane, who put the ONE at the top of his rankings, put it best: "Speed is not the only thing that matters."

For most people: the ThermoWorks ThermoPop 2. SmokingMeatGeeks reports 2 to 3 second reads and 0.1-degree resolution, which is inside the range FSIS gives for thermocouples and plenty for a pork chop. It doesn't fold, which is a smaller drawback than it sounds. Its US list price is around $39, per ThermoWorks pricing tracked through 2026, but check the current price before you buy.

If speed and a folding probe are worth the money: the ThermoWorks Thermapen NANO. The ONE was the benchmark: about one second per The Kitchn's 2026 guide and Food Network, ±0.5°F and IP67 per SmokingMeatGeeks, a backlit display that rotates, a probe that folds flat, and a 5-year warranty. ThermoWorks has since retired the ONE in favor of the NANO, which at $119 on preorder costs more than the $74 closeout Homebrew Finds reported for the ONE. The NANO claims about 0.4 seconds, ±0.5°F from -4°F to 248°F, built-in magnets and a brighter display, according to the ThermoWorks press release and Homebrew Finds. One tradeoff is worth knowing: Homebrew Finds lists the ONE's battery life at 2,000 hours and the NANO's at 1,000. For a home cook, 0.4 seconds against one second is a difference you won't notice over a skillet.

Mid-range and budget folding options: the Lavatools Javelin PRO Duo and the ThermoPro TP19H. Both are folding designs aimed below the Thermapen's price. The published facts available here don't give read times, accuracy or prices for either, so I can't rank them against the ThermoPop 2 on numbers. The one reported finding is a caution: SmokingMeatGeeks saw moisture intrusion in a Javelin Pro Duo after cleaning. If you choose either, run the ice-water check the day it arrives.

The compact drawer thermometer: the OXO Good Grips Chef's Precision Digital Instant Read Thermometer. It is the small, fixed-probe kind you keep with the spatulas. The sources here carry no read time, accuracy or price for it, so treat it as a candidate to check against the same ice test rather than a recommendation on the evidence.

What is worth paying for. Read time up to a point: FSIS's 2 to 5 seconds for a thermocouple is fine, while a 10-second thermistor gets tiresome over a hot pan. A thin tip, because it leaves a smaller hole and reaches the center of a thin chop. A backlit display, which The Kitchn singles out for readability; SmokingMeatGeeks found the Sizzle's display darkened in strong direct sun, so if you grill outdoors, that matters. Water resistance, if you'll wash it under the tap.

What is overkill for weeknight dinners. Sub-second reads. A range up to 572°F, unless you make candy, which The Kitchn notes needs a higher maximum. Accuracy beyond ±2°F, which is Penn State's standard for food safety; ±0.5°F is lovely, but a chicken thigh doesn't care about the difference.