How Accurate Is Apple Watch Calorie Burn?

The Apple Watch estimates calories burned, and the estimate is a rough one. In a 60-person lab test, the Apple Watch 6 got a “poor accuracy” verdict for calories in all five activities, while its heart rate readings were far tighter. A 2020 review of 158 publications found that no brand of wearable was accurate for energy expenditure. Read each burn figure as a range.

How accurate are Apple Watch calories?

Wrist-worn calorie error in the kindest lab result is near 15%, and harsher ones often pass 30%. Study design, activity and the way error is counted all move the number, so the table sets four sources side by side.

Source	Who and what was tested	What it found
Shcherbina et al., J Pers Med 2017	Apple Watch and six other wrist devices, 60 adults, checked against indirect calorimetry	No device got calorie error under 20%. The Apple Watch had the lowest overall error of the seven
Hajj-Boutros et al., Eur J Sport Sci 2023	Apple Watch 6, 60 young adults, five activities, checked against a gas analyzer	Calorie variation of 14.68–24.85%, rated poor in every activity
Fuller et al., JMIR mHealth 2020	Review of 158 publications on nine brands	For energy expenditure, no brand was accurate
Germini et al., JMIR 2022	Review of 65 articles on wrist trackers	Mean absolute percentage error above 30% for all brands

The table uses two error measures. Mean absolute percentage error is the average miss as a share of the true value. The coefficient of variation is how much readings scatter around their own average. By either one, a reading of 400 kcal is a guess with a wide edge.

Both lab tests measured energy from the oxygen people used and the carbon dioxide they breathed out, which is the standard way to do it. The Stanford paper came out in 2017, so its watches are old models. The Apple Watch 6 test is the newer check. It had 30 men and 30 women, about 25 years old on average, with heart rate read against a Polar H10 chest strap and energy against a Metamax 3B gas analyzer, while they sat, walked, ran, lifted weights and cycled.

The Stanford authors set their line for acceptable error at under 5%. Heart rate cleared it in most activities. Calories cleared it in none, on any device. In their energy results the lowest median errors across devices came while walking (31.8%) and running (31.0%), and the highest while sitting (52.4%).

Both studies used healthy adults exercising in a lab. They say nothing about illness, pregnancy or medicines that change heart rate, and your doctor is the person to ask about those. Apple has released newer watches since. I couldn’t open an independent lab test of them, so everything here describes the original Apple Watch and the Apple Watch 6.

Open the Health app and write down your average active calories for a normal week. You’ll use that number below.

Why is Apple Watch heart rate good but calorie burn poor?

Heart rate is read straight from your wrist. Calories are worked out from it with a model of your body. In the Stanford test the Apple Watch had a median heart rate error of 2.0%, and 2.5% while walking. No device got its calorie error under 20%, so the Apple Watch’s was at least ten times larger. At 150 beats per minute, 2.0% is 3 beats. Even the best device in that test, at 27.4% of a 400 kcal day, was 110 kcal off, about a small snack on a food log.

The Apple Watch 6 test points the same way. Its heart rate variation stayed under 5% in every activity, and its calorie variation ran from 14.68% to 24.85%.

Apple’s own page names height, weight, gender and age as inputs to its calorie math. The formula isn’t published, and the Stanford authors call such calculations proprietary. They suspect some now use heart rate, and they note how much body size matters: 10,000 steps have been observed to represent anywhere from 400 to 800 kcal, depending on a person’s height and weight. My own sums for that count are in how many calories 10,000 steps burn.

More sensors do help. A meta-analysis of 60 studies found that adding a heart rate or heat sensor to an accelerometer lowered the error in most activity types. It also found that research-grade devices beat commercial ones for total daily energy but lost to them during walking and sedentary tasks. The sensors have improved, yet the last step, where a model turns body signals into an energy figure, is still the unreliable part.

The 2020 review found that Fitbit, Apple Watch and Samsung measured steps accurately in laboratory settings, and that the Apple Watch and Garmin were the most accurate for heart rate. Those two readings are the ones I’d lean on.

Your own error may not match the average either. The Stanford authors found that device error differed with sex, body mass index and skin tone, so a table of medians describes a crowd, not you.

Check which question the screen answers, too. The Move ring shows active calories burned. That is a smaller number than your whole day’s burn, which also includes what your body spends at rest, and people mix the two up.

What does a 30% error look like in food?

On a day when you truly burn 400 active kcal, an error between 27.4% and 52.4% means the watch could show anything from 190 to 610 kcal. That is a miss of one to two bananas either way. The percentages come from the Stanford study, and the arithmetic is mine: 27.4% was the lowest median any single device reached, 31.8% was the median across devices while walking, and 52.4% was the median while sitting.

Error	Watch could show (kcal)	Gap either way (kcal)	Gap in bananas
27.4%	290 to 510	110	1.1
31.8%	273 to 527	127	1.3
52.4%	190 to 610	210	2.1

Each range runs from 400 kcal times (1 − error) up to 400 kcal times (1 + error), so the gap is 400 × 27.4% = 110 kcal. A banana is 88 kcal per 100 g by the specific Atwater factors, and one banana weighs 115 g, so 88 × 1.15 is 101 kcal. Divide the gap by 101 and you get the last column.

The studies measure how far readings sit from the truth and say nothing about which way, so your watch may run high or low. A 600 kcal day carries half again as much gap: 27.4% there is 164 kcal. A watch that ran 110 kcal high for you every day would add 770 kcal over a week. Real errors rarely line up that neatly. A consistent bias is still the case worth testing for.

To get your own gap, take the average active calories you wrote down earlier. Multiply by 0.274 for the best case. For a walking-heavy week, use 0.318. Any plan that hangs on a smaller difference is one the watch can’t settle.

Which calorie burn tracker is the most accurate?

None is accurate for calories, and the ranking moves with the study. The Apple Watch had the lowest overall error of the seven devices in the Stanford test. In the 2020 review it and Garmin were the most accurate for heart rate, yet no brand was accurate for energy.

Fitbit errs in a known direction. A review of 52 studies, 41 of them pooled in a meta-analysis, found an average underestimate of 2.77 kcal per minute for energy, with population limits from −12.75 to +7.41 kcal per minute. Over a 30-minute session that average is about 83 kcal low.

The 2023 test also compared two other watches. The table shows how widely each one’s readings scattered across the five activities.

Watch	Calorie variation, lowest to highest activity	Heart rate variation, lowest to highest activity
Apple Watch 6	14.68–24.85%	under 5% in all five
Polar Vantage V	16.54–25.78%	2.44–8.80%
Fitbit Sense	13.44–29.66%	4.14–10.76%

On calories, nobody wins: the Fitbit Sense posted both the lowest and the highest figure. The Apple Watch 6 does win on heart rate. The ranges run across activities and aren’t paired, so a row isn’t a head-to-head score. The 2020 reviewers add that devices are redesigned constantly. Evidence like this dates fast.

By the evidence I could open, the most accurate calorie burn tracker isn’t something you can buy. Pick a watch for what it measures well, steps and heart rate, and don’t pay extra for calorie claims. The other watch questions sit under smartwatches.

How can I make Apple Watch calorie burn more accurate?

Calibrate it and keep your personal details current. Apple says calibration is meant to improve distance, pace and calorie readings in Workout and in the Activity app. Check that Location Services and Motion Calibration & Distance are on in your iPhone settings. Then walk or run outdoors for about 20 minutes at your usual pace, on a flat, open route with good GPS reception, and repeat at each speed you use.

Calibration won’t close a gap of 25% or more. It teaches the watch your stride. Apple says the learning continues on later outdoor walks and runs. No study I found measured whether calibration cuts calorie error, so treat it as a tune-up. If calibration ever seems to make things worse, you can reset it: Watch app, My Watch, Privacy, Reset Fitness Calibration Data.

Update your height and weight in the Health app too, since the watch uses them for its calories.

To cross-check a walking day, put your step count and weight into the steps to calories calculator and compare the result with the watch. When the two are far apart, trust neither one alone.

The check I trust most takes three weeks. Log your food honestly for three weeks and watch your weight trend. If the trend stays flat, you ate about what you burned, whatever the watch said. That is your real maintenance, and it comes from your own data. One week proves little.

If the watch’s error stays steady, this week against last week still tells you something. The studies don’t say whether the error is steady, so I treat that as an assumption and test it against the weight trend.

Should I add Apple Watch calories to my food goal?

You don’t have to. If you do, count only part of it. With an error that starts above a quarter, adding the full reading can turn a 110 kcal gap into a daily habit, and a goal from a calculator already assumes some activity level, so the watch can end up counted twice. Whether Apple Health can log food at all is a separate question, answered in does Apple Health have a food tracker.

Calsa reads active energy from Apple Health but leaves it out of your goal until you switch on “Count activity from Health”. Then it adds half of the active energy to what’s left for the day, and your weight shows as a smoothed trend you can check against the watch. There is no Apple Watch app yet, so everything from the watch arrives through Apple Health.

If a doctor has set a calorie target for you, keep it, and ask whether the watch has any role in it.

Use the watch to see whether this week was more active than last. What you really burn shows up in your food log and your weight trend. And if you were hoping the watch could count what you eat, calorie counter watch explains why it can’t.