A thermometer that reads five degrees high turns a perfectly rested medium-rare steak into medium and a probe-tender brisket into a dry one. Most people never check theirs. The tests take ten minutes, cost nothing, and tell you whether the number you're cooking to is the number the meat is actually at. This guide covers the two standard reference points, how to correct for altitude, how to test each type of thermometer you own — instant-read, leave-in probe, the grill's lid dial, and an infrared gun — and what to do when one is off. It pairs with our thermometer roundup and wired vs. wireless probe comparison.
Two Reference Points
Ice bath: 32°F
The more reliable of the two because it doesn't depend on altitude. Fill a tall glass with crushed ice — crushed, not cubes — then add cold water just to cover the ice. Stir for 30 seconds and let it sit a minute. Insert the probe into the center of the ice slurry, not touching the glass, at least two inches deep. Wait for the reading to settle. It should read 32°F. Most quality thermometers are within a degree; a reading two or more degrees off is worth correcting.
Boiling water: 212°F at sea level
Bring a pot of water to a rolling boil. Insert the probe into the water without touching the bottom or sides of the pot and wait for the reading to settle. At sea level it should read 212°F — but water's boiling point drops with altitude, roughly 1°F for every 500 feet, so at 5,000 feet water boils around 203°F. Look up your elevation and use the corrected number. Barometric pressure also shifts it a degree or so, which is why the ice bath is the primary check.
| Elevation | Water boils at (approx.) |
|---|---|
| Sea level | 212°F |
| 1,000 ft | 210°F |
| 2,000 ft | 208°F |
| 3,000 ft | 206°F |
| 5,000 ft | 203°F |
| 7,000 ft | 199°F |
| 10,000 ft | 193°F |
Test at both ends: a thermometer that's accurate in ice and off in boiling water has a slope error, and it's the higher end that matters for barbecue.
Testing Each Type
Instant-read thermometers
Thermocouple instant-reads (the fast, expensive ones) are usually accurate within a degree and rarely drift. Thermistor instant-reads (slower, cheaper) can drift a few degrees over time. Test both in the ice bath every couple of months and any time one takes a fall. Many models have a calibration or offset function in the settings; some cheaper ones have a small nut under the dial that you turn with a wrench while the probe sits in ice water.
Leave-in probes and wireless probes
Probes for smoker controllers and wireless thermometers live in hot, greasy environments and get their cables kinked, and they drift more than instant-reads. Test each probe individually — a four-probe system can have one bad probe. Most base units and apps have a per-probe offset setting; if a probe is more than three or four degrees off in the ice bath and can't be corrected, replace it. Wireless probes with sealed batteries test the same way; a probe that reads erratically or jumps around in the ice bath has moisture inside and is done. See our wireless probe roundup.
The grill's lid thermometer
Assume it's wrong. Lid dials measure the air at the top of the dome, which can be 30–50°F different from the grate where the food sits, and the bimetal coils inside them are cheap and drift. Test by pulling the dial out (most are held by a nut) and holding the stem in boiling water; adjust the calibration nut on the back if it has one. Better: ignore it and put a grate-level probe on a clip next to the food.
Infrared thermometers
An IR gun reads surface temperature and depends on the surface's emissivity — how well it radiates heat. Most dark, matte surfaces (a pizza stone, a cast-iron pan) are close to the default emissivity setting on the gun; shiny stainless is not and will read wildly low. To check an IR gun, aim it at the surface of an ice bath (should read close to 32°F) or at a black pot of boiling water (close to boiling point). Don't try to check it against a probe thermometer in the same pot; they measure different things. Our pizza oven guide covers using an IR gun on the stone.
Correcting an Off Thermometer
- Offset in settings. If the device supports it, enter the difference. A probe reading 34°F in ice gets a −2°F offset.
- Mechanical adjustment. Dial thermometers with a calibration nut: hold the probe in ice water and turn the nut until the needle reads 32°F.
- Mental correction. If neither is available and the error is consistent (same at both ends), write the offset on a piece of tape on the thermometer and subtract it. Do this only as a stopgap.
- Replace. A thermometer that's off by different amounts at the two reference points, or that gives different readings on repeated tests, can't be trusted at any setting.
Probe Placement Errors That Look Like Calibration Errors
Before you blame the instrument, check technique. Common false readings:
- Touching bone. Bone conducts heat differently; readings jump. Aim for the thickest part of the muscle, away from bone.
- In a fat seam. Fat lags the muscle temperature. Move the probe.
- Too shallow. The sensing element on most probes is in the first half inch to inch of the tip; it needs to be fully inside the meat, and on thin cuts, inserted from the side.
- Probe cable near the fire. A cable routed over the firebox reads hot. Route it away from direct heat.
- Not waiting for the reading to settle. Instant-reads take 2–5 seconds; thermistors longer. Watch for the number to stop moving.
How Often to Test
Every couple of months for thermometers in regular use, after any drop or immersion, before a big cook, and whenever a reading seems wrong compared to the visual and touch cues you've learned. A quick ice bath while the smoker comes up to temperature is a good habit.
Bottom Line
Ice bath first, boiling water second with an altitude correction, every probe individually, and the lid dial last if at all. Offset what you can, replace what you can't, and keep one thermometer you've verified as the reference for the others.
Frequently Asked Questions
How do I calibrate a meat thermometer at home?
Make an ice slurry of crushed ice and just enough cold water to cover it, insert the probe into the center without touching the glass, and wait for it to settle — it should read 32°F. Then test in boiling water, correcting for your altitude. Use the device's offset setting or calibration nut to fix small errors.
What temperature should boiling water read on a thermometer?
212°F at sea level, dropping roughly 1°F per 500 feet of elevation — about 203°F at 5,000 feet. Look up your elevation before judging the reading.
Is the thermometer on my grill lid accurate?
Usually not. It measures air at the top of the dome, which can be 30–50°F off from grate level, and the bimetal coil drifts. Use a grate-level probe clipped next to the food instead.
How often should I check my thermometer's accuracy?
Every couple of months in regular use, after any drop or immersion, and before a big cook. A quick ice bath while the smoker heats up takes two minutes.
Can an infrared thermometer be calibrated?
You can check it by aiming at the surface of an ice bath or a black pot of boiling water. Accuracy also depends on the emissivity setting — shiny surfaces read low. Don't compare an IR gun to a probe in the same pot; they measure different things.
Affiliate disclosure: The Smoke Signal earns a commission on qualifying purchases made through Amazon and eBay links on this page, at no extra cost to you. Pricing is shown as tiers ($ / $$ / $$$) because street prices move constantly — check the retailer for current numbers. We do not publish invented specs or reviews.