How to Use a Meat Thermometer on the Grill or Smoker: Where the Probe Goes, Cut by Cut

A meat thermometer answers two different questions, and most bad advice comes from mixing them up. One is is it safe, and the USDA’s Food Safety and Inspection Service (FSIS) answers that with a short chart of minimum internal temperatures. The other is is it done, and on a grill or smoker the answer is usually a cook’s texture target well above the safety number: a brisket at 145°F is safe and still as tough as a boot. This guide keeps the two apart, then covers where the probe goes in each cut, how deep each kind of thermometer has to go, where the pit probe belongs, and how to check that the thermometer is telling the truth.

Nothing here comes from our own testing. The placement rules are the makers’ and FSIS’s, and the doneness targets are named cooks’ and testers’, each attributed where it appears.

The two numbers: safe and done

FSIS publishes the safe minimums, and its chart opens with the condition that makes them work: “Cook all food to these minimum internal temperatures, as measured with a food thermometer, before removing food from the heat source.”

The high smoker numbers are about collagen melting into gelatin, not about safety, which is why they sit so far above the chart. Going the other way is where the labels matter. Some testers pull chicken breast below 165°F and rely on holding it at temperature for a set time (ThermoWorks holds 157°F for 31 seconds). That is a time-at-temperature method that FSIS’s consumer chart does not offer, so 165°F stays the number to cook poultry to unless you are deliberately following one of those methods.

Which thermometer for which job

There are three kinds, and each has a different minimum depth. FSIS’s table gives the floor for each type.

• Instant-read (folding probe or pen). For spot checks and anything thin. Thermocouple types read in 2 to 5 seconds and need about a quarter inch in the food; FSIS says the older dial type averages along 2 to 3 inches of its stem, which is why a dial cannot read a thin chop.
• Wired leave-in (a base unit with probes on cables). For long cooks in a smoker or oven: a meat probe stays in the thickest part and an air probe reads the pit. FSIS’s minimum for an oven-cord probe is half an inch.
• Truly wireless in-meat (the probe is the whole thermometer). FSIS’s minimum for a wireless or smart probe is at least an inch into the thickest part, but every maker marks its own line, and that line is the one to follow (see below).

FSIS also says a thermometer should not stay in food while it cooks unless the maker says it is built for that. A folding instant-read left in a brisket is a melted instant-read.

Where the probe goes, cut by cut

The rule under every row is FSIS’s: “The food thermometer should be placed in the thickest part of the food, away from bone, fat or gristle. For thin foods, the food thermometer should be inserted through the side until it reaches the center of the food.” Bone throws a reading off, and where sources give a direction it reads high, so keep the tip a quarter to half an inch away from it (America’s Test Kitchen’s figure for thighs is at least ¼ in; AmazingRibs keeps a pork-butt probe at least ½ in from the bone).

Whole poultry gets three checks. FSIS: “the innermost part of the thigh, the innermost part of the wing and the thickest part of the breast. Avoid the bone.” A leave-in probe in the breast tells you when to start checking; the instant-read in all three places tells you when it’s done.

Big or uneven cuts get several checks. FSIS says to check a large or irregular roast in several places, and ThermoWorks notes that a big roast or turkey can vary by twenty or thirty degrees from one spot to another. Smoked meat is only as done as its coolest part.

Wireless probes: the line on the probe is the rule

A truly wireless probe keeps its electronics in the pointed half, which the meat keeps at or below the boiling point of water, and its ambient sensor in the handle. Push it in short of the maker’s line and the electronics cook; push the handle into the meat and the ambient reading is meaningless. Every maker marks the line, and they describe it differently:

Two practical consequences follow from those depths. A cut thinner than the probe’s insertion depth cannot be read properly with it, which is why Typhur’s staff send anything under 66.5 mm to an instant-read and MEATER does not recommend its probes for small cuts. And the handle’s heat rating is model-specific: MEATER’s original, Plus and SE probes are rated to 527°F ambient, its Pro probes to 1,000°F “even over open flame”. A probe rated for a 250°F smoker is not automatically rated for a sear. Combustion also asks you to insert its probe once, at the start: pulling it out and repositioning mid-cook confuses its prediction.

Wired probes and the pit probe

A wired probe’s weak point is the cable, not the tip. ThermoWorks rates its Pro-Series cables to 700°F and the joint where the cable meets the probe to 644°F, and FireBoard rates its thermistor probes to 572°F. ThermoWorks’ routing rules for a smoker:

• Use the grommet or probe port if the cooker has one.
• Never feed cables down the chimney: exhaust heat concentrates there and can run hotter than the cooking temperature.
• With no port, close the lid lightly on the cable rather than crimping it.
• Keep the cable off the grates, and if it must cross a hot zone, wrap only that section in a ball of foil. Wrapping the whole length can channel hot air down to the joint.

ThermoWorks also says that, as a general rule, it does not recommend leave-in probes for grilling, because flare-ups can exceed 1,000°F. On a hot grill, use an instant-read.

The pit probe goes where the food is. The dial on a lid sits several inches above the grate and reads a different temperature from the food. Both ThermoWorks and AmazingRibs put the air probe at grate height, about an inch above the grate, with its tip off the metal. They differ on distance from the meat: ThermoWorks clips it about an inch away, and AmazingRibs about 2 inches, because cold meat sits in a pocket of cooler air. Neither cites a measurement, so anywhere an inch or two beside the meat is a defensible choice. Keep it off the edge of the grate and away from the firebox side of an offset.

A wireless probe’s handle reading is not a good stand-in for a pit probe when a fan controller is running the fire. Typhur says the cold steel shaft conducts the meat’s low temperature up to its handle’s ambient sensor early in a cook, which lowers the reading. Our inference: a controller holding the pit to a number that reads low would run the fire hotter than you set it.

How to read an instant-read

Every source we read agrees on the method: push the probe past where you think the centre is, then draw it back slowly and watch for the lowest reading. That low point is the thermal centre, and it is the number that counts. ThermoWorks’ worked example: readings of 136, 128, 123 and 127°F on the way out mean the centre is at 123°F. Kenji López-Alt’s reason for pulling rather than pushing is control: a probe pushed in tends to jump forward, while one drawn out moves smoothly.

On a thin steak, a chop or a burger, go in from the side. FSIS suggests lifting the piece onto a clean spatula or plate if the grill is too hot for your fingers. Check more than one spot as the meat nears its target; ThermoWorks finds steaks run hotter near the ends.

Carry-over: how much it rises after you pull it

Meat keeps cooking after it leaves the heat, and the amount depends on how hot the cooker was and how big the cut is. ThermoWorks measured it once for each item, in ovens, on an uncovered metal tray:

On a low-and-slow smoker the rise is small: AmazingRibs puts it at about 5°F at 225°F, and up to 20°F for a roast at 350°F. Foil or a cooler adds to it. So a reverse-seared steak or a hot-roasted pork loin can come off 5 to 10 degrees early; a brisket coming off at 203°F will barely move. Poultry is the exception: FSIS measures its 165°F before the food leaves the heat, and America’s Test Kitchen advises against pulling poultry early.

Check that the thermometer is right

The ice bath. FSIS’s method is a large glass of ice cubes topped with water and stirred, with the probe at least 2 inches in, touching neither the sides nor the bottom; it should read 32°F. ThermoWorks’ version is stricter, and it is the one that tells you whether a thermometer is actually off: fill the glass to the top with ice (crushed is best), add water to about half an inch below the top of the ice, let it sit a minute or two, and stir the probe gently in the middle. Made that way, “your ice bath will be 32.0°F within ±0.1°F”; a cup of water with a few cubes in it “can be 12 or more degrees too high.” If the ice floats, pour off water and add ice, because the water below it is not at 32°F.

Boiling water reads 212°F at sea level only. Colorado State University Extension’s table puts the boiling point at about 208°F at 2,000 ft, 203°F at 5,000 ft, 198°F at 7,500 ft and 193°F at 10,000 ft, so check against your own elevation.

How close is close enough. FSIS says most food thermometers read within 2 to 4°F; Texas A&M AgriLife Extension treats ±2°F as accurate. A digital thermometer comes with a stated accuracy, and ThermoWorks’ rule is not to adjust one that reads within its own specification. If yours reads 2°F high and cannot be adjusted, add 2°F to every target.

Where to next

The site’s grill and smoker showdowns are where these probes earn their keep: the offset smokers under $700, kamado grills and electric smokers all run long enough that a leave-in probe and a pit probe pay for themselves. For the outdoor-kitchen side of the same cook, see Outdoor Kitchen Layout.