How Accurate Is Apple Watch? Calories, Blood Oxygen, VO2 Max

By Eric Parker, founder of TrueCal. Published . Sources checked .

Written with AI assistance from the cited studies. Every claim was checked against its source. Editorial policy.

Key takeaways

  • Apple Watch heart rate is accurate, with pooled studies finding an average bias of 0.27 beats per minute.13
  • Calorie estimates are much less accurate. In a Stanford test of seven wrist devices, none estimated calorie burn with less than 20 percent error, and across studies the error ran high in some and low in others.161819
  • Apple says its Blood Oxygen readings are for general fitness and wellness, and in a controlled low-oxygen test the watch read high.210
  • Against direct lab tests, Apple Watch VO2 max estimates ran about 13 percent low.1112

How accurate is Apple Watch overall?

Apple Watch is accurate for heart rate but much less accurate for calories burned. In a Stanford test of seven wrist devices, none estimated calorie burn with less than 20 percent error, though the Apple Watch did best. Its blood oxygen reading usually lands within about 4 points of a blood test, and its VO2 max estimate ran about 13 percent low against lab tests.161311

A 2026 review pooled 82 studies of the Apple Watch and found that accuracy depends heavily on what is being measured.13

MeasurementWhat studies foundBest use
Heart rateAverage bias of 0.27 beats per minute, with most readings within about 7 beats of the referenceTrust it day to day13
Calories burnedError inconsistent and frequently large, in the review's words, from about 10 percent while running to over 150 percent while walkingA rough trend only13
Blood oxygenAverage bias near zero, but single readings ranged about 4 points either side of the referenceA wellness check, not a medical reading13
VO2 maxAbout 6 milliliters per kilogram per minute low on average, around 13 percent errorA long-term fitness trend13

Are Apple Watch calories accurate?

Not very. The watch measures heart rate well, then estimates calories from heart rate, motion and your profile. The estimate is where the error comes in.13

Most of these tests used older watch models. The 2026 review noted that five of its eight calorie studies tested Series 2 or older, so newer models have less evidence behind them.13

  • In a test of 22 healthy adults, four wrist devices including the Apple Watch had 1 to 9 percent error for heart rate but 9 to 43 percent error for energy burned.19
  • In Stanford's test of 60 people with a range of ages, sizes and skin tones, no device got calorie error below 20 percent. Errors were larger during walking and in people with a higher BMI.16
  • A review of 158 studies found Apple Watch among the most accurate brands for heart rate, but in the authors' words, for energy expenditure, no brand was accurate.9

Does Apple Watch overestimate calories?

In several studies, yes. In others it read low. The error can go in either direction, and you can't tell which way it runs for you from the watch alone.1851519

The practical point is the same either way. A watch calorie number is a reasonable way to compare today with last week. It is not precise enough to plan your food around to the calorie.1851519

StudyWho was testedDirection of error
Sun 202311 young men, Apple Watch Series 6, runningTended to read high, up to 53 percent18
Falter 201940 people with heart disease, cycling testRead high5
Nuss 201930 young adults, treadmillHigh for women, low for men15
Wallen 201622 healthy adults, lab protocolRead low19

How accurate is Apple Watch blood oxygen?

In Apple's words, its Blood Oxygen readings are not intended for medical use and are only designed for general fitness and wellness purposes. It is supported on Series 6 or later and all Ultra models. On some US models sold after January 18, 2024, the analysis runs on the paired iPhone and results appear in the Health app.2

Here is what testing against arterial blood, the reference standard, has found.13310

  • Pooled across 9 studies and 969 people, the average bias was close to zero, but individual readings spread from about 4 points low to 4 points high. In people with low oxygen, that spread widened to roughly 8 to 9 points either way.13
  • In 167 emergency patients with a COPD flare, Apple Watch Series 6 averaged within half a point of blood gas results, with 95 percent of differences falling within about 5 to 6 points either way.3
  • In a Duke study that lowered oxygen levels under controlled conditions, both the Apple Watch Series 7 and a Masimo finger oximeter read high. When true saturation was below 88 percent, the watch's average error was 5.82 points, against 3.52 for the finger device. The finger device's readings were also more likely to fall within an accepted error range, 49 percent against 32 percent for the watch.10

How does that compare with a medical pulse oximeter?

FDA guidance lists typical accuracy for cleared oximeters of 3.0 points for finger clip sensors and 3.5 points for reflectance sensors, the kind that sit flat on the skin, across readings from 70 percent to 100 percent. In the Duke low-oxygen test, the watch's error was well above both figures.610

The direction matters too. When a device reads high at low oxygen levels, a reassuring number is the one you can trust least. The FDA advises people using an oximeter at home not to rely on it alone and to contact a health care provider if they are worried about a reading or their symptoms are serious or getting worse.810

Does skin tone affect blood oxygen readings?

For finger oximeters in hospitals, yes. Among readings of 92 to 96 percent, true oxygen was below 88 percent in 17.0 percent of measurements in Black patients versus 6.2 percent in White patients across 178 hospitals. The FDA lists skin pigmentation, poor circulation, skin temperature, tobacco use and nail polish among factors that affect readings. For the Apple Watch itself, the evidence is thin. The Duke study said its limited skin tone diversity restricts what its results can show.17810

How accurate is Apple Watch VO2 max?

Apple's 2021 validation paper reported an average error of 1.4 milliliters per kilogram per minute, with a standard deviation of 4.7, in 221 people held out for testing. Its reference values were projected from submaximal exercise tests, not measured during a maximal test.1

Independent tests that measured VO2 max directly found larger errors.11124

The Caserman study found the watch tended to read high in less fit people and low in very fit people.4

StudyWatchPeopleResult against a lab test
Lambe 2025Mostly Series 9 and Ultra 2306.07 milliliters per kilogram per minute low on average, 13.3 percent error11
Lambe 2026Series 10406.25 milliliters per kilogram per minute low, 13.2 percent error, fitness group right for 14 percent of people12
Caserman 2024Series 71941.4 on the watch versus 45.9 in the lab4

When does Apple Watch update VO2 max?

Estimates come from outdoor walks, runs or hikes on fairly flat ground with good GPS and heart rate signal. Apple's estimates cover 14 to 60 milliliters per kilogram per minute. People taking heart rate limiting medicines, such as beta blockers, can record that in the Health app so the estimate accounts for it.1

In Apple's own data, a typical person's estimates varied by about 1.2 milliliters per kilogram per minute, and 1 in 10 people varied by 2.6 or more. A change of a point or two from one month to the next is within that normal spread.1

Why does VO2 max matter?

In 122,007 adults who had treadmill tests, the least fit quarter had about 5 times the adjusted risk of death of the fittest 2 percent over a median 8.4 years. The gap between below average and above average fitness, a hazard ratio of 1.41, was similar to smoking at 1.41 and diabetes at 1.40. This was an observational study using treadmill results, not watch estimates. It shows fitness tracks with risk. It doesn't show what any one person's number change will do.14

How to track it

Use the watch for what it measures well, and check its calorie number against your own data.

Logging meals in TrueCal and weighing in regularly gives you both halves of this check. Our TDEE calculator gives a starting estimate to compare against.

  • Log what you eat for two to three weeks. Consistency matters more than perfect accuracy.
  • Weigh yourself regularly and watch the trend line. Single days swing with water and food in your gut.20
  • Compare the two. If your weight trend holds steady at your average logged intake, that intake is close to your real daily burn, whatever the watch says. If the trend moves, your burn is above or below it.21
  • Treat the watch's active calories as a trend. A higher number than last week likely means you moved more. It doesn't tell you the exact amount.16

Common myths

  • Myth, my watch tells me exactly how many calories I burned. Lab studies show errors of 20 percent or more are common, in both directions.1613
  • Myth, a normal blood oxygen reading on my watch rules out a problem. At low oxygen levels the watch tends to read high, so a normal-looking number can hide a real drop.10
  • Myth, my VO2 max fell one point, so I lost fitness. Apple's own data shows typical estimates vary by about that much.1

A note on medical advice

This article is for general information and isn't medical advice. Talk with your clinician about your own situation.

Worked example

In the Caserman study, VO2 max averaged 45.9 in the lab and 41.4 on the watch. 45.9 minus 41.4 is 4.5 milliliters per kilogram per minute, and 4.5 divided by 45.9 is about 9.8 percent below the lab value.4

A closer look

In a Duke study that lowered oxygen levels under controlled conditions, the Apple Watch Series 7 read high. When true saturation was below 88 percent, the watch's average error was 5.82 points, against 3.52 for a Masimo finger oximeter, so at low oxygen a reassuring number is the one you can trust least.10

Questions and answers

Is Apple Watch more accurate than other fitness trackers?

For heart rate and calories it tends to rank near the top. In the Stanford test it had the lowest overall error of seven devices, but its calorie error still exceeded 20 percent.16

Is the Apple Watch heart rate accurate?

Yes, in most conditions. Pooled studies found an average bias under 1 beat per minute.13

Is Apple Watch blood oxygen FDA approved?

No. Apple describes it as a wellness feature not intended for medical use, and the FDA does not review oximeters sold as general wellness products before they go on sale.27

Why doesn't my treadmill run update my VO2 max?

Apple Watch estimates VO2 max from outdoor walks, runs and hikes with GPS.1

Should I use Apple Watch calories to set my food intake?

Use them as a rough guide. Your logged intake and weight trend over a few weeks give a more reliable picture of what you burn.169

Keep a food log with TrueCal

Type a meal in plain words and TrueCal adds it to your log, so your intake and your weight trend sit side by side.

Get TrueCal on the App Store Get TrueCal on Google Play

Sources

  1. Using Apple Watch to Estimate Cardio Fitness with VO2 max. Apple. 2021.
  2. Blood Oxygen app on Apple Watch. Apple. 2026.
  3. Ir J Med Sci. Arslan. 2024.
  4. JMIR Biomed Eng. Caserman. 2024.
  5. JMIR Mhealth Uhealth. Falter. 2019.
  6. Pulse Oximeters Premarket Notification Submissions guidance. FDA. 2013.
  7. FDA Proposes Updated Recommendations to Help Improve Performance of Pulse Oximeters Across Skin Tones. FDA. 2025.
  8. Pulse Oximeters. FDA. 2026.
  9. JMIR Mhealth Uhealth. Fuller. 2020.
  10. JMIR Form Res. Jiang. 2026.
  11. PLoS One. Lambe. 2025.
  12. Mayo Clin Proc Digit Health. Lambe. 2026.
  13. NPJ Digit Med. Lambe. 2026.
  14. JAMA Netw Open. Mandsager. 2018.
  15. Am J Health Behav. Nuss. 2019.
  16. J Pers Med. Shcherbina. 2017.
  17. N Engl J Med. Sjoding. 2020.
  18. J Sci Med Sport. Sun. 2023.
  19. PLoS One. Wallen. 2016.
  20. Physiological Reports. Bhutani. 2017.
  21. Lancet. Hall. 2011.