Apple Watch VO2 Max Accuracy: What the Research Shows

Apple Watch VO2 max estimates tend to underestimate true cardiorespiratory fitness by a meaningful margin. Two recent validation studies, each comparing the watch against gold-standard lab testing, found that it missed the mark by roughly 6 to 6.25 mL/kg/min on average, with a mean absolute percentage error hovering around 13%. That is a noticeable gap if you care about the exact number, but the picture gets more interesting when you look at how the watch performs for tracking changes over time versus pinpointing your absolute fitness level.

How the Apple Watch Generates a VO2 Max Estimate

The Apple Watch does not directly measure how much oxygen your body consumes. Instead, it infers VO2 max (which Apple calls “Cardio Fitness”) from your heart rate response during outdoor walking, running, or hiking on relatively flat ground. The algorithm requires a few conditions to be met: adequate GPS signal, a solid heart rate reading from the wrist sensor, and an increase of roughly 30% above your resting heart rate during the activity.¹ If any of those conditions fall short, the watch simply won’t produce an estimate for that session.

The underlying logic is straightforward in concept: fitter people have lower heart rates at a given exercise intensity. By combining your walking or running pace (tracked via GPS) with your heart rate during that effort, the watch builds a model of how hard your cardiovascular system works to sustain a given workload. Apple layers in user-provided data like age, sex, height, and weight to refine the prediction. But the watch never sees your actual oxygen consumption, so it is always estimating from indirect signals.

What the Lab Comparisons Found

The gold standard for measuring VO2 max is a graded exercise test in a laboratory, where you breathe through a mask that directly analyzes inhaled and exhaled gases while running on a treadmill or pedaling a stationary bike.² Two independent validation studies compared Apple Watch estimates against this kind of testing, and their results line up remarkably closely.

A 2025 study with 28 participants wearing recent Apple Watch models found the watch underestimated VO2 max by an average of about 6.1 mL/kg/min, with a mean absolute percentage error of roughly 13.3%.³ A separate study published in Mayo Clinic Proceedings: Digital Health, specifically testing the Apple Watch Series 10, reported nearly identical results: an average underestimation of about 6.25 mL/kg/min and a mean absolute percentage error of 13.2%.⁴

The consistency between those two studies is telling: the watch reliably reads low by several units, and the spread of individual errors is wide. In the Mayo Clinic study, the limits of agreement ranged from roughly minus 18 to plus 6 mL/kg/min.⁴ That means some individuals saw their watch reading land close to their true value, while others were off by a substantial amount in either direction. The average tells one story; individual accuracy tells another.

Why a 13% Error Matters (and When It Doesn’t)

To put that error in perspective, consider someone whose true VO2 max is 45 mL/kg/min. A 13% error could place their Apple Watch reading anywhere from about 39 to 51. That range spans what most classification charts would call “fair” to “excellent” depending on your age and sex. If you are using the number to compare yourself against a fitness table or to satisfy a medical threshold, the watch’s estimate is too imprecise to be trusted on its own.

But most people checking their Apple Watch aren’t trying to hit a clinical benchmark. They want to know whether their fitness is going up, staying flat, or declining over months of training. Here the research offers some encouragement. In a study of patients with congenital heart disease who underwent two separate lab tests, the smartwatch-estimated changes in peak VO2 between tests correlated strongly with the lab-measured changes. For every 1 mL/kg/min shift detected in the lab, the watch detected about 0.9 mL/kg/min of change.⁵ The absolute number was off (the watch tended to read a few units higher than lab values in that population), but it tracked real shifts reliably. If you treat the number as a trend indicator rather than a ground truth, it becomes considerably more useful.

Where the Errors Come From

Several factors conspire to push the Apple Watch’s estimate away from your true VO2 max, and understanding them helps explain why some people get readings closer to reality than others.

The most fundamental issue is that the watch relies on optical heart rate sensing, a technology called photoplethysmography. Green LEDs shine into the skin and measure how much light is absorbed by blood pulsing through capillaries. This works well at rest and during steady-state exercise, but accuracy drops during activities with irregular movement patterns.⁶ Wrist movement during vigorous running, trail running, or anything involving lots of arm swing can create motion artifacts that corrupt the heart rate signal. Ambient light can add another layer of noise, especially outdoors.⁷

Because the VO2 max algorithm depends entirely on the heart rate data being accurate, any heart rate error feeds directly into the fitness estimate. If the watch overcounts your heart rate during a brisk walk, it thinks your cardiovascular system is working harder than it actually is, which pushes the VO2 max estimate down. This likely explains the consistent underestimation pattern seen across studies: the watch may be slightly overreading heart rate at exercise intensities, or its algorithm may be calibrated conservatively.

Environmental conditions add another variable. Heat, humidity, altitude, caffeine intake, dehydration, and sleep debt all raise heart rate independently of fitness, and the Apple Watch has no direct way to separate those influences from your actual cardiorespiratory capacity. Research modeling personalized heart rate responses to exercise found that when environmental and contextual factors were accounted for, prediction errors dropped to about 5% of average VO2 max, roughly half the error seen in standard Apple Watch estimates.⁸ The watch’s current algorithm doesn’t appear to fully correct for those variables, which contributes to day-to-day fluctuations in your readings even when your underlying fitness hasn’t changed.

Skin Pigmentation and Sensor Accuracy

Optical heart rate sensors have a known limitation with darker skin tones: melanin absorbs more of the green light used by the sensor, which can weaken the signal. This is a real concern, and the research paints a nuanced picture for Apple Watch specifically.

A 2024 study testing the Apple Watch Series 9 found that heart rate measurements were more variable for participants with darker skin pigmentations compared to lighter ones.⁹ A separate 2025 study examining multiple wearable brands found that skin pigmentation did not significantly predict heart rate error magnitude for the Apple Watch in particular, though it did for another brand tested. However, the study noted that missing data and outliers were disproportionately observed among participants with darker skin across all devices tested, including Apple Watch.¹⁰

What does this mean practically? If you have darker skin, your Apple Watch heart rate readings might be slightly noisier during exercise, and the device may occasionally fail to produce a reading at all. Since VO2 max estimation depends on accurate heart rate data, that noise can carry over into less reliable fitness estimates. Apple has not publicly addressed this in its algorithm documentation. Wearing the watch snugly and keeping the sensor clean helps, but the underlying physics of optical sensing through melanin-rich skin is a hardware-level constraint that software can only partially compensate for.

How Apple Watch Stacks Up Against Other Smartwatches

The Apple Watch is not alone in producing VO2 max estimates with double-digit percentage errors. One validation study comparing multiple devices noted that Polar watches showed a mean absolute percentage error above 10% (specifically around 13.2%), and Fitbit devices hovered around 10%. Garmin watches, which use algorithms developed by Firstbeat Technologies, consistently came in below 10% error, making them the most accurate of the major consumer brands for this specific measurement.¹¹ An older study of wrist-worn trackers from 2019 similarly found that most devices exceeded 10% error, with only the Garmin Forerunner 920XT falling below that threshold at about 7.3%.¹²

That said, the Apple Watch performs well on other health metrics. A large systematic review and meta-analysis found that for heart rate and blood oxygen saturation, the Apple Watch showed stronger agreement with reference measures than Garmin, Fitbit, and Withings devices.¹³ The takeaway is that Apple’s hardware is capable of strong physiological measurement, but its VO2 max algorithm specifically seems to lag behind Garmin’s Firstbeat-powered approach. Whether that gap narrows with future software updates remains to be seen.

People with Heart Conditions

Researchers have been particularly interested in whether smartwatch VO2 max estimates are useful for people with cardiac conditions, since cardiorespiratory fitness is a powerful predictor of outcomes in that population. Low VO2 max is associated with increased risk of cardiovascular events and all-cause mortality, making it a clinically meaningful number to track.¹⁴

The results from a study of patients with complex congenital heart disease were cautiously promising. The watch’s estimates correlated well with lab-measured peak VO2, though the watch consistently read about 3.1 mL/kg/min higher than the lab in that group.⁵ More encouragingly, changes between two testing sessions tracked closely. For patients who improved or declined between lab visits, the watch picked up on those shifts with strong fidelity.

However, there is a significant caveat. Heart rate measurement accuracy is lower in individuals with arrhythmia, including atrial fibrillation.⁶ Since an irregular heartbeat produces an irregular optical signal, the watch’s heart rate algorithm can struggle, and the cascading effect on VO2 max estimation could be substantial. If you have an arrhythmia, your Apple Watch VO2 max number should be treated with extra skepticism. The watch might still detect broad trends in fitness over months, but individual readings on any given day may be unreliable.

Getting Better Readings from Your Watch

You cannot fix the fundamental limitations of wrist-based optical sensing, but you can minimize the factors that degrade accuracy. Since the algorithm requires outdoor walking, running, or hiking on relatively flat terrain, indoor treadmill sessions won’t generate a new estimate (no GPS). Hilly routes may also be excluded if the incline exceeds the threshold the algorithm accepts.

A few practical considerations can help reduce noise in your readings:

Fit and placement: Wear the watch snugly enough that it doesn’t slide around on your wrist but not so tight that it restricts blood flow. The sensor needs consistent skin contact to get a clean optical signal.
Consistent conditions: If you want to compare readings over time, try to record them under similar conditions. Running in 95-degree heat will temporarily inflate your heart rate and deflate your VO2 max reading compared to the same run at 65 degrees. Morning caffeine, poor sleep, and dehydration do the same.
Sufficient duration: Longer workouts give the algorithm more data to work with. Short bursts of activity may not meet the minimum requirements for generating an estimate.
Watch the trend line: The Health app shows your Cardio Fitness over weeks and months. A single data point is noisy; the rolling average over many sessions is far more informative.

Apple’s algorithm also updates your VO2 max passively during walks, not just tracked workouts. This is a double-edged sword. On one hand, it means you accumulate more data points without having to remember to start a workout. On the other, a casual stroll in the heat after a heavy meal may produce a reading that drags your average down in ways that don’t reflect your actual fitness.

The Overestimation and Underestimation Split

An interesting inconsistency across studies deserves attention. The two large validation studies in healthy adults both found the Apple Watch underestimates VO2 max, by about 6 mL/kg/min on average.³ Yet the study in congenital heart disease patients found the opposite: the watch overestimated peak VO2 by about 3.1 mL/kg/min.⁵

This discrepancy likely reflects how the algorithm was trained. Apple’s model was almost certainly built on data from healthy adults performing standard activities. Individuals at the lower end of the fitness spectrum, including people with cardiac conditions, may have heart rate patterns that the algorithm interprets through a “healthy person” lens, leading it to overpredict their capacity. Meanwhile, very fit individuals who push high VO2 max values in the lab may see their estimates pulled toward the middle by an algorithm that hasn’t been exposed to enough elite-level data. Both ends of the fitness distribution get compressed toward the center, which is a common failure mode for prediction models trained on average populations.

This compression effect matters for exactly the people who might benefit most from tracking: those with low fitness due to chronic conditions, and athletes training at high intensities. Both groups should expect their Apple Watch readings to be directionally informative but potentially biased in opposite ways.

What VO2 Max Actually Tells You About Health

The reason all this accuracy discussion matters is that VO2 max is one of the strongest predictors of long-term health outcomes. Low cardiorespiratory fitness is associated with higher risk of dying from cardiovascular disease and from all causes.¹⁴ Apple explicitly ties its Cardio Fitness feature to this research, classifying users into “low,” “below average,” “above average,” and “high” categories and flagging low cardio fitness as a health concern.

The catch is that those classification boundaries are narrow. Apple’s “low” Cardio Fitness threshold varies by age and sex, but the cutoffs are typically separated by just a few mL/kg/min. When your watch has a demonstrated error of 6 mL/kg/min on average, and potentially much more for any individual reading, a classification of “below average” versus “above average” is not trustworthy from a single estimate. Over many readings across months, the average might settle closer to your true category, but the uncertainty window should make anyone cautious about treating a single Apple Watch classification as a health alarm.

Researchers investigating wearable-based VO2 max have noted that the technology’s value lies less in replacing a lab test and more in providing continuous, low-barrier monitoring that can flag trends and motivate behavior change. A person who sees their Cardio Fitness trend steadily downward over six months has a real signal worth discussing with their doctor, even if the individual numbers are imprecise. The trend is the feature; the specific number is the limitation.