Apple Watch calories burned: how accurate is the count?

You finish an hour on the bike, the red ring closes and the watch says 650 kcal. The number looks precise down to the last calorie, which makes it tempting to take it at face value. Researchers who have compared the Apple Watch with indirect calorimetry, the reference method that works out energy use from the oxygen you breathe in and the carbon dioxide you breathe out, urge caution.

The watch does not measure the energy you burn directly. It estimates it from what its sensors pick up (movement, heart rate, GPS outdoors) and the information you have given it. Apple explains that the watch selects the most appropriate inputs for each workout: on a treadmill, for example, it also relies on the accelerometer.

The result: the Apple Watch is one of the best watches for heart rate, but energy expenditure remains its weak point, as it is for every competitor. The most complete review so far, a 2025 meta-analysis of 56 studies published in the journal Physiological Measurement, found an average error of 27.96% for energy expenditure, compared with 4.43% for heart rate and 8.17% for steps. Every subgroup analysed for energy expenditure was above the 10% threshold researchers use to call a measurement valid.

This is the question people ask most: why does the watch show two such different numbers? The red Move ring in the Activity app shows only active calories, the ones you burn by moving, walking, climbing stairs or exercising. Those are what your daily goal rewards.

Resting calories are, in Apple’s developer documentation, the energy your body uses to keep its normal resting state: breathing, circulating blood, maintaining cells. This is your basal metabolic rate, and it is usually the biggest share of a day’s energy use. No sensor measures it: Apple says the watch uses your height, weight, sex and age to calculate how many calories you burn.

Total calories add the two together. Apple Health actually stores Active Energy and Resting Energy as two separate categories, and the documentation states that active energy must not include the resting energy burned over the same period. A workout summary can also show total calories: they include what you would have burned anyway during that time.

An average error of 28% does not mean your watch is off by 28% every time. It can be close on a steady run and drift during a game of tennis. The 2020 meta-analysis of 60 studies by O’Driscoll and colleagues confirms it: the accuracy of wrist and arm devices depends on the type of activity, and adding heart rate to movement data reduces the error in most cases.

The person matters too. In the Stanford study, the watches’ error was higher for men, for people with a higher BMI, for people with darker skin and during walking. Two people of the same weight running side by side at the same pace can therefore get estimates of different quality.

You won’t turn your Apple Watch into a lab, but a few settings prevent the biggest errors. These are the ones Apple recommends in its own documentation, and together they take less than half an hour.

If you track your meals, the temptation is obvious: the watch says you burned 500 kcal, so you allow yourself 500 kcal more. The problem is that the watch’s error then goes straight onto your plate. On 500 kcal, a 28% error is about 140 kcal, and if it is an overestimate, part of the calorie deficit you thought you were creating disappears.

On the other hand, eating none of it back can make your deficit bigger than planned on hard training days. The middle ground is to eat back part of your active calories, for example half, then check what your weight tells you.

Energy balance has two sides, and both are estimates. The burn shown by your watch has a margin of error; so does the intake you log in an app. In the OPEN study, which compared what 484 adults reported eating with their energy use measured by doubly labelled water, men underreported their intake by 12 to 14% and women by 16 to 20% on 24-hour recalls. Apps that estimate a meal from a photo have their own limits, which we cover in our article on how accurate photo calorie counting is.

The good news is that these errors become manageable if you keep the same method from week to week. Your scale takes everything into account: your real burn, your real intake and whatever neither the watch nor the app can see. By comparing your average weight from one week to the next, you know whether your balance is really negative, whatever the red ring says. Weigh yourself under the same conditions, for example in the morning before breakfast, and look at the average rather than a single reading, which can swing with water and digestion.

The Apple Watch measures heart rate well and steps fairly well, but its calorie burn remains an estimate, with an average error close to 28% according to the latest meta-analysis and gaps of more than 100 kcal over an hour of exercise. The Move ring shows only active calories; total calories add a basal metabolic rate calculated from your profile. Keep that profile up to date, calibrate the watch outdoors, tighten the band during exercise and start a matching workout for every session. For your diet, eat back only part of your active calories and judge the result on the trend in your weight. If the scale stops moving despite impressive burn figures, our guide to the weight loss plateau helps you find what is holding you back.