A watch VO2 max is a useful estimate, not a measurement. Across validation studies, the average error is small for some devices, but the error for any one person can be large: often 6 mL/kg/min or more, and in some studies well over 10% of the true value. Apple Watch tends to read low, Garmin tends to read closer but still misses, and both are better at showing your trend than your exact number.
Below is what the research says about each brand, why the numbers drift, how to read a score like 47.7, and when a lab test is the better tool.
TL;DR
- Watches estimate VO2 max from heart rate, pace and other data. They do not measure oxygen use the way a lab gas analyzer does.
- Apple Watch underestimated VO2 max by about 6 mL/kg/min on average in recent studies, with an error of roughly 13% to 16%. Garmin studies show a wider range, from about 3% to 10% error depending on the model and the athlete.
- Individual error is much bigger than average error. A watch can be right on average and still be far off for you.
- Use the watch for trends over weeks and months. Use a lab test when you need a number you can rely on.
Table of Contents
- What your watch is actually estimating
- What the studies say, in one table
- Apple Watch: consistent low readings
- Garmin: closer on average, still variable
- Why the average looks good and your number may not
- What throws the estimate off
- How to read a number like 47.7
- Use the watch for trends, not for exact scores
- When a lab test is worth it
- Perspective: our take on wrist scores
- Find a lab test near you
- FAQ
- Sources
What your watch is actually estimating
VO2 max is the most oxygen your body can take in and use during hard exercise, reported in milliliters of oxygen per kilogram of body weight per minute (mL/kg/min). A lab measures it directly: you run or cycle to exhaustion while a mask captures every breath. That method, indirect calorimetry, is the reference standard.
A watch never sees your breath. It infers fitness from how hard your heart appears to work at a given speed. Apple says its estimate draws on heart-rate response, motion-sensor data, your profile and GPS-derived data. Google Health describes running pace relative to heart rate, compared against benchmarks for your age, sex and weight.
That is a reasonable shortcut, but it rests on assumptions. If your heart rate is higher or lower than the model expects for your fitness, the score moves even though your lungs and muscles haven't changed.
What the studies say, in one table
The numbers below come from peer-reviewed validation studies that compared a watch with laboratory testing. MAPE is the mean absolute percentage error: the average size of the miss, whether high or low.
| Device or group | Result vs. lab | Source |
|---|---|---|
| Apple Watch (2025 study, Ireland) | Underestimated by 6.07 mL/kg/min on average; MAPE 13.31% | PLOS ONE |
| Apple Watch Series 10 (2026) | Mean difference of -6.25 mL/kg/min; MAPE 13.2% | Mayo Clinic Proceedings: Digital Health |
| Apple Watch Series 7 (2024) | MAPE 15.79%; RMSE 8.85 mL/kg/min | Smartwatch accuracy study |
| Garmin fenix 6 (2023) | MAPE 6.85% in athletic participants | Technologies, MDPI |
| Garmin Forerunner 245 (endurance athletes) | Underestimated by about 4.05 to 4.73 mL/kg/min | Forerunner 245 validity study |
| Personally owned devices (2024) | Apple underestimated by 9.7 ± 9.1; Garmin by 5.9 ± 6.8 mL/kg/min | Medicine & Science in Sports & Exercise |
Two things stand out. First, nearly every study finds watches reading low. Second, the plus-or-minus figures are large. A spread of ±9 around an average underestimate of 9.7 means some people in that study were off by close to 19 mL/kg/min, while others were nearly spot on.
Apple Watch: consistent low readings
The 2025 Irish study found Apple Watch underestimated VO2 max by 6.07 mL/kg/min on average. The authors concluded the estimates need further refinement before clinical use. A 2026 study of Series 10 found almost the same gap, so the problem has not gone away with newer hardware.
The 2024 Series 7 study shows what that looks like in practice. Lab VO2 max averaged 45.88 mL/kg/min while the watch averaged 41.37. Reliability was poor (ICC 0.47), which is why a score such as 47.7 should not be treated as exact.
If you are fit and your Apple Watch says you are "below average," this is the most likely explanation. The error is not random: it leans low.
Garmin: closer on average, still variable
Garmin results are more mixed, and they depend on who is wearing the watch. In the fenix 6 study, athletic participants showed acceptable agreement with a 6.85% MAPE. In the Forerunner 245 study, moderately trained athletes had a MAPE of 2.8% to 4.1%, while highly trained athletes had 9.4% to 10.4%.
In other words, the fitter you are, the less the algorithm tends to capture. That pattern shows up in reviews too: wearables can provide valid estimates in healthy untrained adults and recreational athletes, while their usefulness in elite endurance sport is questionable.
There is also a context effect. A 2024 study found two smartwatch models reached excellent agreement during outdoor running, with MAPE of 8.8% and 8.0%. Controlled outdoor runs on a good signal are the best case for any watch.
Why the average looks good and your number may not
The 2022 INTERLIVE systematic review and expert statement is the clearest summary of this. Wearables that use exercise-based algorithms had a near-zero average bias of -0.09 mL/kg/min, which sounds excellent. But the individual limits of agreement ran from about -9.92 to +9.74 mL/kg/min.
That means the errors cancel out across a crowd. Some people read too high, some too low. For you, the question is where you fall in that range, and you cannot know without a lab test.
To put 10 mL/kg/min in perspective, it is roughly the gap between an average 40-year-old and a very fit one. A 10-point miss can change which fitness category you appear to be in.
The same review found that devices using exercise data did better than those using resting data (bias of -0.09 versus 2.17 mL/kg/min). So watches that need a real run or walk to produce an estimate are generally the more trustworthy kind.
What throws the estimate off
Because the watch reads your heart rate, anything that changes heart rate without changing fitness can distort the score. Apple's own documentation lists terrain, gradient, heat, caffeine, alcohol and heart-rate-affecting medications. The 2026 Series 10 study makes the same point, adding individual variation in heart-rate response.
The common culprits:
- Heat and humidity. Your heart rate climbs to cool you, so the watch reads you as less fit.
- Hills. Apple generates estimates from outdoor walks, runs or hikes on relatively flat ground, under 5% incline or decline, with good GPS and heart-rate signal. Hilly routes give the model poor input.
- Caffeine, alcohol, illness, poor sleep. All raise heart rate for a given pace.
- Beta blockers and similar medications. These lower heart rate and can push the estimate the other way. Ask your prescriber how your medication affects heart-rate-based fitness metrics.
- Loose strap or wrist sensor noise. A bad optical signal means bad data in.
- Wrong max heart rate or profile settings. Garmin's model relies on your maximum heart rate. One analysis suggests accuracy falls when runners underestimate it by a wide margin, so check your setting.
- Cycling. Garmin accuracy is generally stronger for running; for cycling the watch needs a power meter to estimate at all.
The Mayo Clinic Proceedings researchers add a broader caution: real-world activity is "uncontrolled", without the standardization of a clinical test, and algorithms and hardware keep changing. A score from your watch this year may not be comparable to the same score after a software update.
How to read a number like 47.7
A reading of 47.7 means the watch estimates your body can use 47.7 milliliters of oxygen per kilogram of body mass per minute. It is not a percentage and not a grade. Apple classifies cardio fitness by age group and sex, so the same 47.7 means different things for a 25-year-old man and a 60-year-old woman.
Now apply the error bars. If you wear an Apple Watch and it says 47.7, the studies suggest your lab value could plausibly be several points higher. If you wear a Garmin, the miss is smaller on average but could go either way. Treat 47.7 as "somewhere in the mid-to-high 40s, possibly higher," not as a precise figure.
For how real treadmill-tested populations are distributed by age and sex, our breakdown of VO2 max norms from 7,783 treadmill tests is a better yardstick than a wrist estimate.
Use the watch for trends, not for exact scores
The most defensible way to use a watch VO2 max is to watch the direction. A consistent device worn the same way will carry its own bias along with it, and that bias partly cancels when you compare you-to-you.
There is a limit, though. A 2022 runner study found an overall smartwatch error of 5.7%. Around a true value of 50 mL/kg/min, a displayed change of at least 2 mL/kg/min was needed before it could confidently represent a real physiological change. So a one-point rise after a good week is noise. A steady climb of 3 or 4 points over a training block is more convincing.
To get the most from the trend:
- Keep the same watch, the same firmware habits and the same profile settings.
- Compare similar conditions: flat route, mild weather, no hard caffeine or alcohol the day before.
- Look at a rolling average over four to eight weeks, not single readings.
- Confirm the big moves, up or down, with a lab test.
Reviewers also call for longitudinal studies because frequent updates may affect data quality over months or years. That is another reason not to read too much into small shifts.
When a lab test is worth it
Laboratory graded exercise testing with gas analysis remains the gold standard. A 2025 review states that wearables offer scalable fitness surveillance but should not replace laboratory measurements when precise risk stratification is required. The same review notes that VO2 max and related fitness measures are increasingly treated as vital signs for guiding risk assessment and monitoring lifestyle changes.
A lab test makes sense when:
- You train seriously. The Garmin data shows error growing in highly trained athletes, exactly the people who need exact zones.
- Your watch score doesn't match how you feel or perform. A very low reading despite solid race times is a common sign of underestimation.
- You want to set training zones. Wrong inputs mean wrong intensity. If you race 10K or longer, see how lactate threshold compares with VO2 max for planning.
- You want a baseline before a training block or a health change, then a repeat afterward to see what actually changed.
- A clinician is asking. If you have symptoms or a medical condition, talk to your doctor about appropriate testing rather than relying on a consumer device.
If you book one, preparation matters. Our guide to resting 24 to 48 hours and matching the protocol explains how to avoid skewing the result. If cost is the hesitation, see what a VO2 max test costs in the U.S..
Perspective: our take on wrist scores
We list DEXA and VO2 max testing clinics for a living, so we hear this question constantly: "My watch says 41, my friend's says 52, who is right?" Our honest answer is that neither number is wrong in a way you can fix by tweaking settings. They are estimates built from heart rate, and heart rate is influenced by far more than fitness.
We think watches are a good thing. They get people moving, and they give a free running log of direction. Where they go wrong is when a single displayed number becomes a verdict: someone decides they are "poor" for their age, or an athlete trains at the wrong intensity because the watch said so.
Our rule of thumb is simple. Let the watch tell you whether you are trending up. Let a lab test tell you where you actually stand. The same logic applies to body composition: a scale or a bioimpedance gadget gives a quick read, while a DEXA scan gives a measurement you can anchor to. If you are building a full picture of your fitness, pairing a lab VO2 max with a body composition scan is more informative than either alone, and it is also a handy way to see why body fat beats BMI.
— Erik
Find a lab test near you
Dexascans helps you search by city or ZIP code, compare cash prices and book directly with the clinic, with no physician referral needed. Listings include VO2 max and resting metabolic rate add-ons alongside DEXA scans, so you can pair tests in one visit where a clinic offers it. If you are in South Florida, start with VO2 max testing in Miami. To add a body composition baseline, see our body composition scan guide.
FAQ
How accurate is VO2 max on smartwatches?
On average, moderately accurate. Across studies, error ranges from about 3% to 16% depending on the device, model and who is tested. Individual differences are large: one systematic review found that a single reading can differ from true VO2 max by nearly 10 mL/kg/min in either direction.
Can I trust the Apple Watch VO2 max?
Trust it as a trend, not as an exact number. Recent studies found Apple Watch underestimated VO2 max by about 6 mL/kg/min on average, with a MAPE of roughly 13%. If you are fit, your true value is probably higher than the watch says.
Is 47.7 a good VO2 max?
It depends on your age and sex, since fitness categories are set by those. A value of 47.7 mL/kg/min is a strong figure for many adults, but because it is a watch estimate it could be off by several points. A lab test would confirm where you actually fall.
Is Garmin more accurate than Apple Watch for VO2 max?
In the studies above, Garmin devices showed smaller average errors than Apple Watch, particularly for running. But results differ by model and by how trained the wearer is, and accuracy drops in highly trained athletes. Neither brand replaces a laboratory test.
Sources
- Validity of Estimating the Maximal Oxygen Consumption by Consumer Wearables (INTERLIVE review)
- Investigating the accuracy of Apple Watch VO2 max (PLOS ONE, 2025)
- Accuracy of VO2 max Estimates From Apple Watch Series 10
- Assessing the Accuracy of Smartwatch-Based Estimation of VO2 max
- Validation of Aerobic Capacity (VO2max) and Lactate Threshold, Garmin fenix 6
- Validity of VO2max estimates from the Forerunner 245
- Accuracy of personally owned wearable activity monitors
- Validation of VO2max and sweat loss metrics from smartwatches
- Accuracy of wearables for determining maximal oxygen uptake (systematic review)
- Assessment of peak oxygen uptake with a smartwatch
- Using Apple Watch to Estimate Cardio Fitness with VO2 max
- Learn about your VO2 max, Google Health Help Center
- Comparative review of VO2max measurement protocols