How to Calibrate a Probe Thermometer

Calibrate a probe thermometer using two reference points: the ice-point method (a slushy ice-water slurry should read 0°C) and the boiling-point method (rolling boiling water should read 100°C at sea level). Australian food businesses handling potentially hazardous food generally must keep a probe thermometer accurate to ±1°C, so check it regularly, record the results, and service or replace any thermometer that drifts outside tolerance. Key takeaways: A probe thermometer used with potentially hazardous food generally must be accurate to ±1°C — the ice-point and boiling-point methods let you verify both ends of the working range.. Ice-point (0°C in an ice-water slurry) is the quicker, safer everyday check; boiling-point (100°C at sea level) confirms accuracy at the hot end and is affected by altitude.. There is no single legislated calibration interval — set a risk-based frequency (many kitchens check weekly to monthly, plus after any drop or extreme use) and follow the manufacturer's advice.. Record every calibration check: date, thermometer ID, reference point, reading and the action taken. Records are what an auditor or EHO wants to see.. If a thermometer reads outside ±1°C and cannot be adjusted, take it out of service and repair or replace it — a wrong reading is worse than no reading.. Interpreting a calibration check: Result, Action. Within ±1°C of the reference — Pass — record and continue. Outside ±1°C but adjustable — Reset the offset or turn the calibration nut, then re-verify. Outside ±1°C and not adjustable — Apply a documented correction factor (temporary) or replace. Erratic or non-repeatable — Remove from service — this cannot be corrected. Why calibration matters and what the Code expects Temperature control is the backbone of food safety, and every cold-holding, hot-holding, cooking, cooling and reheating decision you make rests on one assumption: that your thermometer is telling the truth. A probe that has drifted by 3°C can make dangerous food look safe. That is why Standard 3.2.2 of the Australia New Zealand Food Standards Code generally requires a food business handling potentially hazardous food to have a temperature-measuring device that is readily accessible and accurate to ±1°C. An accuracy of ±1°C means that when the display shows 5°C, the true temperature sits somewhere between 4°C and 6°C. Calibration is how you confirm the device still meets that tolerance. Importantly, the Code sets the accuracy requirement but does not prescribe a specific calibration method or interval — that is left to you to manage sensibly. The ice-point and boiling-point methods described below are the two widely accepted ways Australian food authorities recommend for checking a probe thermometer in a commercial kitchen. Requirements and interpretation can vary between states and territories, so confirm the detail with your local council or state food regulator. Standard 3.2.2 generally requires ±1°C accuracy for businesses handling potentially hazardous food.. Calibration verifies accuracy; it does not, on its own, make an inaccurate probe correct unless the device can be adjusted.. The Code does not mandate a set frequency — you decide based on risk and manufacturer advice.. Ice-point method (checking 0°C) The ice-point method is the everyday workhorse: it is fast, cheap, safe and reliable, and it checks accuracy at the cold end where your fridges, cool rooms and chilled displays operate. Because water freezes at a stable, reproducible 0°C regardless of altitude, it is the method most kitchens use for routine checks. The key to an accurate ice point is a proper slurry — lots of crushed or small ice with just enough cold water to fill the gaps, not a glass of water with ice floating on top. Too much water and the mixture sits above 0°C; too little contact and the probe never equilibrates. Stir well, keep the probe tip suspended in the slurry (not touching the sides or base of the container), and give it time to settle before reading. Fill a container with crushed or small ice, then add clean cold water until the gaps fill — the ice should not float freely.. Stir into a slushy slurry and let it stand for a minute so it stabilises at 0°C.. Insert the probe into the centre of the slurry, avoiding the sides and base; stir gently.. Wait until the reading stabilises (about 30 seconds to a few minutes depending on the probe).. A correctly calibrated thermometer reads 0°C (within ±1°C, i.e. -1°C to 1°C). Repeat a few times to confirm the readings are consistent.. Boiling-point method (checking 100°C) The boiling-point method checks accuracy at the hot end of the range — relevant for cooking, hot holding and reheating verification. Bring a saucepan of water to a continuous rolling boil, immerse the probe tip in the water (keeping it clear of the base and sides of the pan), and let the reading stabilise for a couple of minutes. At sea level, pure water boils at 100°C, so a well-calibrated probe should read 100°C within ±1°C. Two cautions. First, take care — you are working with boiling water and a hot pan, so protect your hands and keep the cord and body of the instrument away from the heat. Second, boiling point falls with altitude and this genuinely matters in parts of Australia. The NSW Food Authority notes that water boils at around 94°C in the highest towns in NSW, which can make a perfectly good thermometer appear to fail the ±1°C tolerance. If you are well above sea level, adjust your expected reading for your elevation (boiling point drops by roughly 1°C for every 285–300 metres of altitude) rather than assuming the thermometer is faulty. Where altitude makes the boiling-point check unreliable, rely on the ice-point method, which is unaffected by elevation. Bring clean water to a rolling boil in a deep saucepan.. Immerse the probe tip in the water, keeping it away from the base and sides.. Let it stabilise (about 2–3 minutes) and read the temperature.. Expect 100°C at sea level (±1°C); reduce the expected value if you are at altitude.. Never leave the probe unattended in boiling water and follow the manufacturer's maximum temperature rating.. Adjusting, correction factors and when to retire a probe After a check, one of three things is true. If the reading is within ±1°C of the reference, the thermometer passes — record it and carry on. If it reads outside tolerance, check whether your device can be adjusted: some digital probes have a reset or offset function and many dial (bimetal) thermometers have a calibration nut under the head that you turn while the probe sits in the reference (ice slurry is easiest and safest). Adjust, then re-check to confirm it now reads correctly. If the device cannot be adjusted, you have two options. You can apply a documented correction factor — for example, if the probe consistently reads 2°C in ice water, you know to subtract 2°C from every reading — but this is error-prone in a busy kitchen and is generally a stop-gap, not a fix. The cleaner option is to take the thermometer out of service and repair or replace it. Because a compliant probe thermometer is generally inexpensive relative to the cost of a food safety incident, replacement is often the sensible call. Any probe that gives erratic, non-repeatable readings should be retired regardless of the average, because inconsistency cannot be corrected. Within ±1°C: pass — record and continue.. Outside ±1°C but adjustable: reset the offset or turn the calibration nut, then re-verify.. Outside ±1°C and not adjustable: apply a documented correction factor as a temporary measure, or replace.. Erratic or non-repeatable: remove from service immediately — this cannot be corrected.. How often to calibrate and setting a frequency The Food Standards Code does not specify how often you must calibrate, so you set a risk-based frequency and document your reasoning. In practice, many Australian kitchens do a quick ice-point check somewhere between weekly and monthly, with the exact interval driven by how heavily the probe is used, how critical its readings are to your food safety controls, and the manufacturer's recommendation. Beyond the routine schedule, calibrate whenever confidence in the reading is in doubt: after a probe has been dropped or knocked; after exposure to extreme heat or cold (for example probing hot oil or being left in a freezer); when readings look suspicious or inconsistent with what you expect; when a new thermometer first arrives; and after a battery change on a digital unit. If your food safety program or third-party audit scheme (for example in aged care, childcare or under a specific standard) prescribes a frequency, follow the stricter requirement. Some businesses also send instruments to an accredited calibration laboratory (for example one with NATA accreditation) periodically for a traceable certificate, keeping in-house checks as the day-to-day control. Set a routine interval (commonly weekly to monthly) based on usage, criticality and manufacturer advice.. Always re-check after a drop, extreme temperature exposure, or any suspicious reading.. Check new thermometers before first use and after battery changes.. Follow the stricter interval where an audit scheme or food safety program specifies one.. Keeping calibration records From an auditor's or environmental health officer's point of view, an uncalibrated-looking thermometer with no records is a red flag even if it happens to be accurate. Standard 3.2.2A introduces additional food safety management and evidence requirements for certain higher-risk businesses, and many food safety programs and audit schemes expect you to demonstrate that your monitoring equipment is reliable — a simple calibration log is how you prove it. Keep it near your temperature records so the whole temperature-control story hangs together. A workable log captures, for each check: the date; which thermometer (a unique ID or serial number if you have more than one); the method used (ice point or boiling point); the reference temperature expected; the actual reading; whether it passed ±1°C; the action taken (none, adjusted, correction factor applied, or removed from service); and the initials of the person who did it. If you outsource calibration to a contractor, keep their certificate showing the date, the instrument's serial number, the correction factor, and the equipment type and location. Retain records for at least the period your state regulator or food safety program requires. Log date, thermometer ID, method, expected reading, actual reading, pass/fail and action taken.. Store calibration records alongside your routine temperature logs.. Keep contractor calibration certificates (date, serial number, correction factor, equipment details).. Have the log ready to show an EHO or auditor on request.. Worked examples: A boiling-point check at altitude: A kitchen in a high-altitude town runs a boiling-point check and the probe reads about 96°C in rolling boiling water. Because boiling point falls with altitude, the thermometer is not necessarily faulty, since the expected value is lower there. They rely on the ice-point method instead, which is unaffected by elevation. A weekly ice-point check: A cook makes a dense ice-water slurry, suspends the probe in the centre clear of the sides and base, and waits for the reading to settle. It reads 0.5°C, within the ±1°C tolerance, so it passes. They repeat it twice to confirm the readings are consistent, then record the date, reading and result in the calibration log. Checklist: Probe thermometer is accurate to ±1°C and readily accessible in the food handling area. A calibration frequency is set in writing (e.g. weekly–monthly) based on use and manufacturer advice. Ice-point check performed using a proper slurry (ice with just enough water, not floating). Boiling-point check performed where needed, with altitude allowance applied if above sea level. Readings are consistent across repeats, not just correct on average. Any out-of-tolerance probe is adjusted and re-verified, or removed from service. Calibration log records date, thermometer ID, method, reading, pass/fail and action. Contractor calibration certificates (if used) are on file with serial number and correction factor. Probe is cleaned and sanitised before and after calibration and between foods. Staff know to re-check a thermometer after any drop or extreme temperature exposure. Common mistakes: Using a glass of water with ice floating on top instead of a proper dense slurry — this reads above 0°C and makes a good thermometer look wrong.. Letting the probe tip touch the base or side of the container or pan, so it reads the vessel's temperature rather than the water.. Reading before the display has stabilised — pull the trigger too early and the number is meaningless.. Ignoring altitude on the boiling-point check and condemning a good thermometer that reads around 94–98°C because it 'should' read 100°C.. Checking only one reference point — an ice-point-only check misses drift at the hot end that matters for cooking and hot holding.. Calibrating but keeping no record, so there is nothing to show an auditor or EHO.. Continuing to use a probe that gives erratic, non-repeatable readings by 'averaging' them, instead of retiring it.. Never re-checking after a probe has been dropped, frozen or pushed into hot oil beyond its rating..