An inaccurate thermometer is worse than no thermometer at all — it gives you false confidence. If your probe says 165°F and your chicken is actually at 155°F, you have a food safety problem. If it says 135°F when the actual temperature is 145°F, you’ll overcook your steak chasing a doneness you’ve already passed.

Here’s how to check your thermometer’s accuracy, how to adjust it if it’s off, and when to replace it entirely.

The ice water method

This is the most reliable calibration check for home use. Ice water holds a stable temperature of exactly 32°F (0°C) at sea level regardless of how much ice or water you add, provided ice is still present. This makes it a perfect reference point.

How to do it:

1. Fill a tall glass or container completely with ice cubes. Don’t use ice chips — the cubes hold their shape longer.
2. Add cold tap water until the glass is almost full. The water should reach the top of the ice without overflowing.
3. Stir the mixture for 15 seconds to equilibrate the temperature.
4. Insert the probe into the centre of the ice water mixture, making sure the tip is not touching the bottom or sides of the glass. The sensing area must be submerged.
5. Wait 30 seconds without stirring, then read the temperature.
6. A correctly calibrated thermometer reads 32°F (0°C). An accuracy of ±1–2°F is acceptable for most thermometers.

The boiling water method

Boiling water provides a second reference point, but altitude affects its usefulness. At sea level, water boils at exactly 212°F (100°C). At 5,000 feet (Denver, Colorado), water boils at approximately 202°F. At 7,000 feet (Santa Fe), it’s around 199°F.

If you live at or near sea level, this method is reliable. If you live at high altitude, look up your local boiling point before using this method, or stick to the ice water test.

How to do it:

1. Bring a pot of water to a full, rolling boil.
2. Insert the probe into the boiling water, making sure the tip is submerged but not touching the pot.
3. Wait 30 seconds and read the temperature.
4. At sea level, it should read 212°F (100°C). At your local altitude, adjust the expected reading accordingly.

How to adjust a thermometer that’s off

Dial (bimetallic) thermometers

Traditional dial thermometers have a calibration nut under the dial face. When the probe is submerged in ice water and reading an incorrect temperature, use pliers or the calibration wrench that came with the thermometer to rotate the dial until it reads 32°F. This type of thermometer can drift and should be checked regularly.

Digital instant-read thermometers

Most modern digital thermometers either have a calibration button (which resets the reading to 32°F when pressed while in ice water) or they don’t offer field calibration at all. Check your thermometer’s manual. If there is no calibration function and the thermometer is consistently off by more than 2°F, note the offset and adjust mentally — or replace the thermometer.

Wireless/Bluetooth thermometers (MEATER, ThermoPro, etc.)

Wireless probe thermometers like MEATER are factory-calibrated and generally do not offer user calibration. If your MEATER probe is reading noticeably off (more than 2°F in ice water), contact the manufacturer. These probes use silicon nitride ceramic tips and precision sensors; if the calibration has drifted significantly, the probe may need replacement. MEATER’s app does include a probe check feature that can flag sensors that have drifted out of spec. For my tested picks across the MEATER range and competitors, see my best Bluetooth meat thermometer guide.

When to replace instead of calibrate

Some signs indicate a thermometer has failed and should be replaced rather than calibrated:

• Inconsistent readings: If you get a different number each time you test in the same ice water without moving the probe, the sensor has failed.
• Off by more than 5°F consistently: A thermometer that far out of spec may have a damaged sensor.
• Physical damage: A bent probe, cracked housing, or water inside the display indicates the unit is compromised.
• After a drop: Dropping a thermometer — especially onto a hard surface — can damage the sensor or displace the bimetallic spring in dial thermometers.
• After extremely high heat exposure: Probe tips damaged by flame or excessive ambient heat (such as placement too close to a grill burner) can produce permanently inaccurate readings.

A good meat thermometer costs between $15 and $100. The cost of a foodborne illness — or a ruined $80 brisket — is significantly higher. Replace a suspect thermometer.

How often should you calibrate?

For instant-read digital thermometers: check calibration every few months, or whenever you suspect an inaccurate reading. These hold calibration well unless dropped or subjected to extreme conditions.

For dial thermometers: check before any important cook. Dial thermometers drift more easily and the calibration nut can shift during transport or handling.

For wireless probe thermometers: check once a season, or if your cooked food consistently seems over- or under-done relative to the probe readings. Test with ice water exactly as described above. See my full meat thermometer buying guide for how this compares across the whole category.

Related guides

Frequently Asked Questions

How do I know if my meat thermometer is accurate?
The ice water test is the standard check: fill a glass with ice, add cold water, stir, and insert the probe. After 30 seconds, a correctly calibrated thermometer reads 32F (0C). If it reads within 1-2F of this, the thermometer is accurate enough for cooking. If it’s off by more than 2-3F consistently, it needs calibration or replacement.

Can I calibrate a digital meat thermometer?
It depends on the model. Some digital thermometers have a dedicated calibration button or setting — check your manual. Others, including most wireless Bluetooth thermometers like MEATER, are factory-calibrated and don’t offer field adjustment. If a non-adjustable digital thermometer is consistently off by more than 2F, note the offset and adjust mentally, or replace the unit.

How do I calibrate a dial meat thermometer?
Submerge the dial thermometer probe in ice water for 30 seconds. While still submerged, use pliers or the calibration wrench (usually included) to grip the hex nut beneath the dial face and rotate it until the dial reads 32F. This resets the bimetallic spring to the correct reference point. Verify by removing and reinserting the probe.

Does dropping a thermometer affect its accuracy?
Yes, significantly. Dropping a dial thermometer can displace the bimetallic spring and cause it to read incorrectly. Dropping a digital thermometer can damage the temperature sensor or loosen internal connections. Always check calibration after dropping a thermometer before trusting its readings for cooking.

A meat thermometer is the single most reliable way to know whether meat is safe to eat and cooked to the texture you want. Timing and colour are unreliable guides. Temperature is not. This guide gives you the numbers you actually need, for every protein, with the context to use them confidently. If you need a thermometer recommendation, my tested guide to the best Bluetooth meat thermometers covers the top five picks.

Internal temperatures are measured at the thickest part of the meat, away from bone, fat, and the pan. Bone conducts heat differently from muscle; measuring near bone gives a false reading.

Safe internal temperatures by protein

Beef doneness guide

The USDA safe minimum for beef whole cuts is 145°F, but most cooks target lower internal temperatures based on preferred doneness. Medium-rare beef at 130–135°F is safe for whole cuts because the surface (where bacteria live) reaches much higher temperatures during searing. Ground beef is different — bacteria are mixed throughout, which is why 160°F is required.

Pull the meat from the heat 5–10°F below your target. Carry-over cooking (the temperature rise during resting) accounts for the difference. Thick cuts (over 1.5 inches) carry over more than thin ones.

Pork and lamb doneness

The USDA updated its pork guidelines in 2011 to align with beef: 145°F with a 3-minute rest. This means properly cooked pork can be slightly pink in the centre — something many home cooks were incorrectly taught to avoid. The pink colour at 145°F is safe. Overcooked, grey, dry pork is not a sign of safety; it is a sign of overcooking.

Fish and seafood

Fish is done at 145°F per USDA guidelines, but the practical test is whether the flesh is opaque and flakes easily with a fork. Fish overcooks quickly — the window between underdone and overdone is narrow. A thermometer is more valuable for fish than for almost any other protein because the visual cues are unreliable until it is already too late.

Low-and-slow: smoking temperatures

Smoking and low-and-slow cooking target much higher final temperatures than standard roasting. The goal is collagen breakdown — turning tough connective tissue into gelatin — which requires sustained heat above 190°F for an extended period. This is why brisket, pulled pork, and ribs are cooked well beyond the 145°F safety minimum.

Common mistakes

Checking temperature too early. The internal temperature rises slowly at first, then stalls (especially in large cuts and brisket — this is the ‘stall’ at around 160–170°F), then rises again. Check at the expected time but don’t panic if it’s lower than expected.

Measuring near bone. Bone conducts heat differently. Always measure in the deepest part of the muscle, at least half an inch from any bone.

Not accounting for carry-over. Pull meat earlier than your target. A 2-inch thick steak can rise 10°F during a 5-minute rest. A large roast can rise 15°F or more.

Using a thermometer with a slow response time. A thermometer that takes 20 seconds to register gives you a reading that reflects where the temperature was, not where it is. Faster is better for thin cuts. For long cooks and roasts, response time matters less.

Testing in one spot only. Temperature varies across a piece of meat. Check the thickest point and avoid the centre of any cavity. For large birds, check both the breast and the innermost thigh — they finish at different times. See my full meat thermometer buying guide for how this compares across the whole category.

Related guides

Frequently Asked Questions

What temperature should chicken be cooked to?
Chicken must reach 165F (74C) internal temperature — both white and dark meat. This applies to whole birds, pieces, and ground chicken. There is no leeway: unlike beef, chicken must hit this temperature for food safety.

Is 145F really safe for pork?
Yes. The USDA updated its pork guidelines in 2011. Whole cuts of pork (chops, loin, tenderloin) are safe at 145F followed by a 3-minute rest. A slight pink colour at the centre is normal and safe at this temperature. Only ground pork requires 160F.

What is carry-over cooking and how much does temperature rise?
Carry-over cooking is the continued rise in internal temperature after meat is removed from the heat. The rise depends on size and thickness: a thin steak might rise 5F, a thick roast can rise 10-15F, and a large brisket can rise more during a long rest. Always pull meat before your target temperature to account for this.

What is the stall in smoking and why does it happen?
The stall is a period during smoking when the internal temperature of a large cut (typically brisket or pork shoulder) plateaus at around 155-170F for several hours before rising again. It is caused by evaporative cooling — moisture evaporating from the meat’s surface cools it as fast as the smoker heats it. It is normal and will pass. Wrapping the meat in butcher paper or foil (the Texas Crutch) shortens the stall by reducing evaporation.

Do I need to rest meat after cooking?
Yes, for most cuts. Resting allows the muscle fibres to relax and reabsorb juices that were pushed to the centre during cooking. A steak benefits from 3-5 minutes of rest; a roast from 10-20 minutes; a whole bird from 20-30 minutes. Cutting too early releases these juices onto the cutting board instead of keeping them in the meat.

Once your roast is at temperature and has rested, a meat slicer makes uniform carving far easier than a knife — especially for thin deli-style cuts of brisket, turkey breast, or standing rib roast. See my best meat slicers guide for home options at every budget.

This is one of the most searched questions about wireless thermometers, and it’s a fair one. You’re trusting a wireless device to tell you whether a piece of meat is safe to eat and cooked to the texture you want. The question of whether to trust it matters.

The short answer is: modern wireless meat thermometers are accurate. The longer answer explains why, what affects that accuracy, and how to get the best readings from any probe thermometer. Every model I reference in this article is tested in depth in my best Bluetooth meat thermometer roundup.

Published accuracy specs from major brands

All of these are well within the margin needed for practical cooking accuracy. Chicken at 165°F is not meaningfully different from chicken at 163°F or 167°F in terms of either safety or eating quality. The ±1°F specification is more than precise enough.

What actually affects accuracy in practice

Probe placement

This is the biggest variable — not the thermometer itself. The internal temperature of a thick piece of meat varies by as much as 20–30°F between the centre and the edge. Insert the probe into the thickest part, avoiding bone, fat pockets, and the cooking surface. A correctly placed probe from a mediocre thermometer will give a better reading than a poorly placed probe from a premium one.

Bone proximity

Bone conducts heat at a different rate than muscle. A probe tip resting against or near bone will read higher than the surrounding muscle temperature. Always keep the probe tip at least half an inch from any bone.

Fat pockets

Intramuscular fat melts at lower temperatures than muscle proteins cook. A probe in a fat pocket reads the fat temperature, not the muscle. In marbled cuts (brisket, ribeye), take multiple readings at different insertion points.

Ambient temperature measurement (for wireless probes with dual sensors)

Probes like MEATER measure both internal meat temperature and ambient oven/grill temperature simultaneously. The ambient sensor is exposed to direct flame and radiant heat, which can affect its accuracy if the probe is placed too close to a burner or hot spot. The internal sensor (at the probe tip, inside the meat) is insulated from this and remains accurate throughout.

Probe condition

A damaged probe tip — bent, corroded, or exposed to extreme direct flame — can read inaccurately. Check your probe before each use. Stainless steel probes (ThermoPro, Inkbird, Weber) are highly durable. Ceramic-tipped probes (MEATER) are hard and heat-resistant but can crack if dropped on a hard surface.

Independent test results

Third-party testing of major wireless thermometers consistently confirms manufacturer accuracy specs. MEATER probes tested in controlled ice water and boiling water conditions have performed within ±1°C. ThermoPro and Inkbird units tested the same way have also come within their published specs in the majority of units reviewed.

Where wireless thermometers do occasionally fall short is in Bluetooth range rather than temperature accuracy. A probe that’s dropped its connection and is showing a cached (stale) temperature reading isn’t inaccurate — it’s not reading. This is a connectivity issue, not a sensor issue, and modern thermometers have significantly improved in this area.

The practical conclusion from real-world testing: if your wireless thermometer tells you your chicken is at 165°F, it is very likely at 165°F ±2°F. That is accurate enough for safe, consistent cooking.

When a wireless thermometer might give a misleading reading

Carry-over cooking during rest: The probe reading continues to rise after the meat comes off the heat. This is real physics, not inaccuracy — the meat is genuinely getting hotter. Pull your meat before your target temperature.

Stall (in smoking): During the barbecue stall, the internal temperature can plateau for hours and even temporarily drop slightly as evaporative cooling counters heat gain. The thermometer is accurately measuring what’s happening.

Reading before equilibration: Insert the probe and wait 5–10 seconds before reading. Instant-read thermometers reach their displayed temperature quickly, but probe thermometers that have been cold can take a few seconds to equilibrate to the surrounding meat temperature.

Cheap or damaged units: Budget thermometers with no published accuracy spec and no brand reputation behind them are where inconsistency is genuinely found. Buying from a reputable brand with a stated accuracy specification is the most reliable way to ensure accuracy. See my full meat thermometer buying guide for how this compares across the whole category.

Related guides

Frequently Asked Questions

Are wireless meat thermometers as accurate as wired ones?
Yes. The accuracy of a meat thermometer depends on its sensor quality and calibration, not whether it transmits wirelessly. Premium wireless thermometers (MEATER, ThermoPro) achieve +/-1C accuracy — the same specification as the best wired probes. The wireless transmission carries the temperature data; it does not affect how the temperature is measured.

How accurate is the MEATER thermometer?
MEATER publishes a +/-1C (+/-1.8F) accuracy spec for all its probes. Independent tests confirm this. The dual-sensor design measures both internal meat temperature and ambient cooking temperature simultaneously, and both sensors perform within this specification under normal cooking conditions.

Can I trust a meat thermometer for food safety decisions?
Yes, if it’s from a reputable brand with a published accuracy spec. USDA temperature guidelines assume a thermometer accuracy of +/-2F. Quality wireless thermometers from brands like MEATER, ThermoPro, and Inkbird all achieve +/-1F or better, making them reliable tools for food safety decisions.

Why is my wireless thermometer giving different readings than I expect?
The most common causes are probe placement (measuring too close to bone or fat), reading before the probe has equilibrated (wait 5-10 seconds after insertion), or carry-over cooking (temperature continues rising after leaving the heat). The thermometer itself is rarely the cause of unexpected readings in quality units.