Athletes: Wearable VO2max Varies ±9.83%, How to Test and Use It

Consumer wearables estimate VO2max, they don’t measure it directly, and that difference matters more than most training plans admit. Exercise-based estimates tend to run closer to lab values, with a headline random error near ±9.83%, while resting-based estimates from HRV or resting heart rate carry a wider margin around ±15.24%. Use the number to track your own trend over weeks and months, not as a clinical result you compare against someone else’s watch.

TL;DR:

• Exercise-based VO2max estimates from wearables are generally more accurate, with a bias close to zero and limits of agreement around ±10 mL/kg/min.
• Resting-based methods tend to overestimate VO2max and have wider variability, making them less reliable for tracking true fitness changes.
• Single VO2max readings can vary widely within their margin of error, so trends over multiple tests are much more meaningful than individual numbers.
• Using a chest strap for heart rate measurement, testing on a consistent route, and entering a lab-tested maximum heart rate improve estimate reliability.
• Wearable estimates are best for tracking general progress, but not for clinical diagnosis or precise assessment, especially in highly trained athletes.

Table of Contents

• Why wearables estimate VO2max rather than measure it
• Exercise-based vs resting-based estimation: which one should you trust?
• What independent studies and meta-analyses actually show
• Why fitter athletes often get worse wearable estimates
• How wearables calculate VO2max behind the scenes
• A practical protocol to get a more reliable estimate
• Reading your VO2max number without overreacting to it
• How Heala turns a noisy VO2max reading into something useful
• What actually matters when you’re chasing this number
• Where to read the original research
• Get more from your VO2max data with Heala
• Sources
• FAQ

Why wearables estimate VO2max rather than measure it

VO2max, in its true form, is the maximum rate at which your body can take in and use oxygen during intense exercise. The gold standard for measuring it is cardiopulmonary exercise testing, known as CPET, which requires you to breathe through a mask connected to a gas analyser while running or cycling to exhaustion on a graded protocol. The equipment tracks the oxygen you inhale against the carbon dioxide you exhale, giving a direct physiological reading rather than a guess.

Your smartwatch has none of that. It cannot analyse your breath, so it builds a proxy instead, a statistical estimate stitched together from whatever data it can collect from your wrist or chest. That proxy is useful, but it’s a different kind of number from the one a exercise physiologist would get you in a lab.

Wearables typically draw on a handful of inputs to build this estimate:

• Heart rate, either from a wrist-based photoplethysmography (PPG) sensor or a chest strap
• GPS-derived pace and distance during outdoor runs
• Heart rate variability (HRV), particularly in resting-based models
• Your entered profile data: age, weight, sex and sometimes resting heart rate

The algorithm combines these inputs with assumptions about how heart rate typically rises with effort in a general population, then outputs a single VO2max figure. It’s a clever piece of modelling, but it’s built on averages, and averages don’t always fit any one individual particularly well. For a fuller breakdown of how your watch turns these inputs into a single displayed number, see our explainer on what that VO2max figure actually represents.

Exercise-based vs resting-based estimation: which one should you trust?

Not all wearable VO2max algorithms work the same way, and the method behind your number changes how much you should trust it. Broadly, devices fall into two families: those that estimate VO2max from data collected during exercise, and those that estimate it from data collected while you’re resting.

Exercise-based algorithms, the approach used by most running-focused GPS watches, calculate VO2max from the relationship between your heart rate and your pace during a submaximal run. Because they’re grounded in an actual physiological response to load, they tend to track closer to laboratory values. Resting-based models, more common in general fitness trackers, infer VO2max from heart rate variability or resting heart rate alone, without you needing to do anything active. That convenience comes at a cost to accuracy.

The INTERLIVE Network’s meta-analysis puts numbers on that gap. Exercise-based estimators showed a pooled bias close to zero, around −0.09 mL/kg/min, with limits of agreement near ±9.83 mL/kg/min. Resting-based methods showed a larger bias, roughly +2.17 mL/kg/min, and wider limits of agreement near ±15.24 mL/kg/min, meaning individual readings can swing considerably further from the true value.

A few practical points follow from that comparison:

• Exercise-based estimates are the more defensible choice if your watch offers a setting for it.
• Resting-based estimates can still be tracked over time, but expect more noise from one reading to the next.
• Neither method should be read as a substitute for a lab test when the number itself, not just its direction, matters to you.

What independent studies and meta-analyses actually show

Peer-reviewed validation work on wearable VO2max has grown considerably, and the picture it paints is consistent: exercise-based models are reasonably close on average but noisy for any single reading, while resting-based models carry a real bias.

The clearest summary comes from the INTERLIVE Network’s systematic review, which pooled data across multiple consumer devices and found that exercise-based algorithms had almost no systematic bias but wide individual scatter, while resting-based algorithms overestimated VO2max on average with even wider scatter. A related systematic review published in 2024 reached a similar conclusion, reporting roughly ±9.83% random error for exercise-based estimators against roughly ±15.24% for resting-based ones when measured against laboratory gas-exchange testing.

Device-specific studies fill in the detail:

• A 2025 validation study on the Apple Watch found it tended to underestimate VO2max, with a mean difference of 6.07 mL/kg/min and a mean absolute percentage error (MAPE) of 13.31% against indirect calorimetry.
• A 2025 validation study on the Garmin fēnix 6 found a MAPE near 7.05% and a Lin’s concordance correlation of around 0.73 under controlled, submaximal outdoor runs with averaged data windows, suggesting reasonable accuracy for recreational athletes under the right conditions.
• A validation study on the Google Pixel Watch 3 reported a low mean bias of −0.53 mL/kg/min but wide limits of agreement, from −12.56 to 11.50 mL/kg/min, and a MAPE of 8.65%, useful for spotting population-level patterns but not precise enough for an individual clinical call.

One figure worth sitting with: the Apple Watch validation study reported a MAPE of 13.31%, meaning a device reading of 45 mL/kg/min could plausibly correspond to a true value anywhere from roughly 39 to 51. That’s the difference between “decent aerobic fitness” and “well above average” on some classification charts, from one device reading alone.

For an individual athlete, this is what limits of agreement mean in practice: they define a range within which most true values are likely to fall around a single device reading, not a guarantee of precision. A watch reading that jumps from 48 to 44 one week isn’t necessarily telling you that your fitness collapsed. It might be well within the device’s own noise band. This is precisely why a single reading matters less than the shape of your trend across many of them.

Why fitter athletes often get worse wearable estimates

It sounds backwards, but well-trained athletes frequently see less reliable wearable VO2max numbers than recreational exercisers, and the reasons come down to how these algorithms were built.

Most consumer algorithms are trained and calibrated on general population data, which means the heart rate-to-effort relationship they assume fits an average person reasonably well but fits an elite endurance athlete poorly. Highly trained runners and cyclists often have unusually efficient running economy, a smaller range between resting and maximal heart rate, and a flatter heart rate response to submaximal pace increases than the model expects. When your physiology sits at the edge of what the algorithm was built to predict, the output drifts.

A few specific failure points show up repeatedly:

• Devices frequently default to an age-predicted maximum heart rate, and a Frontiers systematic review found that an HRmax error of around 15 beats per minute can shift the VO2max estimate by 7 to 9%.
• Superior running economy means a trained athlete can hold a fast pace at a heart rate the algorithm associates with a much slower runner, confusing the fitness calculation.
• A narrow heart rate range between easy and hard efforts gives the algorithm less signal to work with, which increases estimate instability run to run.

The practical upshot is that if you’re already quite fit, treat your device’s number with an extra degree of scepticism, and don’t be surprised if it under-reports relative to how you actually perform.

Pro Tip: If your watch allows you to manually enter a lab-tested or field-tested maximum heart rate rather than relying on the age-based default, do it. It’s one of the few inputs you can correct yourself, and it directly affects the estimate’s accuracy.

How wearables calculate VO2max behind the scenes

Every wearable VO2max figure is built from a small set of sensors feeding an algorithm that most manufacturers keep proprietary. Understanding the pieces helps you understand where the errors creep in.

On the sensor side, most watches rely on wrist-based PPG to read heart rate, which shines light through the skin and measures blood flow changes. Some pair with a chest strap for a more direct electrical reading. GPS supplies pace and distance outdoors, while an accelerometer estimates cadence and, on some devices, effort indoors where GPS isn’t available. All of that sits on top of your entered profile: age, sex, weight, and sometimes a manually entered resting or maximum heart rate.

Each of those components has a known weak point:

• PPG accuracy drops during high-intensity efforts. Motion, sweat and how tightly the watch sits on your wrist all introduce noise, and the INTERLIVE expert statement notes that pressure changes or wrist movement during interval training can create artificial heart rate spikes that distort the resulting VO2max calculation.
• GPS drift affects pace-based inputs. Tree cover, tall buildings and tunnels can all throw off distance and pace readings, which feeds directly into an exercise-based estimate.
• Chest straps generally outperform wrist sensors for heart rate precision, particularly during intervals or hard efforts, because they read the heart’s electrical signal rather than inferring it from blood flow.

Once the sensor data is collected, the algorithm goes to work, and this is largely a black box. Exercise-based models typically fit a curve to the relationship between your heart rate and your running pace across a session, extrapolating what your oxygen consumption would be at a maximal effort. Some newer devices layer in machine learning models trained on larger population datasets, which can improve average accuracy but make it harder for anyone, including researchers, to say exactly why a given reading came out the way it did. The INTERLIVE Network has specifically flagged this opacity as a barrier to comparing devices fairly, since two watches can use entirely different logic to arrive at superficially similar numbers.

A practical protocol to get a more reliable estimate

You can’t turn your watch into a lab, but you can reduce a meaningful amount of the noise in its VO2max readings by controlling what’s within your reach: your setup, your route and your consistency.

1. Pick a flat, familiar outdoor route for testing runs. A consistent 10 to 15 minute submaximal run on the same terrain removes variables like hills and turns that confuse pace-based algorithms.
2. Warm up for five to ten minutes before your test effort. A cold start produces an erratic early heart rate response that can throw off the curve the algorithm is trying to fit.
3. Wear a chest strap if you own one. It reads your heart’s electrical activity directly, avoiding the motion and pressure artefacts that affect wrist-based PPG during faster efforts, as the INTERLIVE Network has documented.
4. Fit your watch snugly, two finger-widths above your wrist bone, if you’re relying on wrist-based sensing instead of a strap.
5. Enter your real profile data, including a field-tested or lab-tested maximum heart rate where you have one, rather than leaving the age-based default in place.
6. Run the same test periodically rather than once, and look at the trend across sessions instead of any single result. The Garmin fēnix 6 validation study found accuracy improved when data was averaged across 15 to 30 second windows and when runs were repeated, which reduces the random error inherent in any one reading.
7. Sync your device regularly so the app has a complete data history to work from rather than gaps that force it to fall back on stale assumptions.

Consistency does more for the reliability of your number than any single piece of hardware. If your test conditions (route, effort, time of day, warm-up) stay the same, the week-to-week movement in your VO2max estimate becomes far more meaningful, even if the absolute number itself carries a wide margin of error.

Pro Tip: Log your test runs using GPS-based cardio tracking so pace, route and heart rate are captured together in one place, making it easier to spot whether a change in your VO2max estimate lines up with a genuine change in effort or output. Heala’s GPS cardio tracking is built for exactly this kind of repeatable outdoor test.

Reading your VO2max number without overreacting to it

The single most useful mental shift with wearable VO2max is to stop treating it as a fixed fact and start treating it as one line on a longer chart. A number that moves from 46 to 47 over a month tells you very little on its own. A number that trends upward across three months of consistent training tells you something real, even if the exact figures aren’t precise.

A few rules keep this grounded:

• Never compare your VO2max number against a friend’s number from a different brand of watch. Different algorithms, different sensors, different assumptions: the figures aren’t built to be cross-compared.
• Look for change that clearly exceeds your device’s own error margin, rather than reacting to small week-to-week wobbles that likely reflect noise rather than fitness change.
• Treat a single alarming reading as a prompt to retest under your standard protocol, not as a verdict.

A meaningful reference point: epidemiological research treats a change of around 3.5 mL/kg/min in VO2max as clinically relevant to long-term health outcomes, yet that figure sits well inside the ±9.83 mL/kg/min margin reported for even the better-performing exercise-based wearable estimates. That gap is exactly why a single reading shouldn’t carry much weight on its own.

If you’re chasing marginal performance gains as a competitive athlete, if a clinician needs a number for medical evaluation, or if you’ve seen a persistent, unexplained shift in your device readings that doesn’t match how you feel in training, that’s the point to book an actual CPET with breath-gas analysis rather than lean further on the watch.

How Heala turns a noisy VO2max reading into something useful

A VO2max estimate on its own, sitting in isolation on a watch face, doesn’t tell you much beyond a rough number. Heala centralises the wearable data behind that number, whether it comes from Fitbit, WHOOP or another connected device, and puts it next to the other signals that give it context.

Rather than showing you a VO2max figure in a vacuum, Heala looks at it alongside your heart rate variability trend, sleep quality and training load over the same window. If your VO2max estimate dips in the same week your HRV drops and your sleep quality declines, that’s a coherent story worth acting on. If it dips on its own with no supporting signal elsewhere, it’s more likely device noise than a real change, and Heala’s daily plans adjust accordingly rather than reacting to a single number in isolation.

Because Heala can pull in multiple runs and readings over time rather than treating each one as a standalone data point, it’s built to flag movement that goes beyond what you’d expect from normal device error, the kind of change worth adjusting a training plan around, and to quiet the noise that isn’t. That’s the difference between a number you glance at and a number you can actually plan around, and it’s the same logic behind how Heala explains VO2max trends rather than just displaying them.

What actually matters when you’re chasing this number

If there’s one thing I’d want every athlete to internalise, it’s that consistency in how you test beats precision in any single device. A cheap watch tested the same way every week will tell you more about your real fitness trajectory than an expensive one used sporadically and compared against someone else’s numbers.

My own priorities, in order: pick one protocol and stick to it, use a chest strap whenever the test actually matters to you, and resist the urge to read anything into a single session. When athletes or the coaches working with them start treating a wearable number as if it were a lab result, that’s when training decisions go wrong, usually by chasing a number rather than the training stimulus that produces real adaptation.

Wearable monitoring earns its place for the week-to-week and month-to-month picture. It has no business replacing a proper CPET when the decision on the table is genuinely high-stakes, whether that’s fine-tuning a competitive athlete’s periodisation or investigating a health concern that needs a clinical answer. For everything in between, a consistent testing habit and a healthy scepticism about the absolute figure will serve you better than any single feature on your watch.

— Kerem

Where to read the original research

The claims in this article draw primarily on the INTERLIVE Network’s systematic review and meta-analysis, the current benchmark for wearable VO2max validity, alongside the 2024 systematic review that reproduced its error estimates. Device-specific validation work includes studies on the Apple Watch, the Garmin fēnix 6 and the Google Pixel Watch 3, each offering concrete accuracy figures worth reading in full alongside the Frontiers review on heart rate defaults.

Get more from your VO2max data with Heala

Reading a VO2max estimate correctly takes more than a single number on a watch face, it takes context from your sleep, your training load and your recovery, all viewed together over time. That’s the gap Heala fills.

Heala connects to your existing wearable, whether that’s Fitbit, WHOOP, Oura or another device, and puts your VO2max trend next to your HRV, sleep and training data so you can see whether a change is real or just noise. Instead of juggling separate apps for nutrition, training and recovery, you get one place where all of it lines up, with AI-assisted meal logging and adaptive daily plans built around what your data is actually showing. The free tier covers the basics, and Heala Pro, from £12.99 a month or £59.99 a year, unlocks deeper analytics and unlimited history for readers who want to track these trends properly over the long term.

If you’re ready to stop guessing what a single VO2max reading means, check Heala’s plans and connect your wearable today.

Sources

• Validity of Estimating the Maximal Oxygen Consumption by Consumer Wearables: A Systematic Review with Meta-analysis and Expert Statement of the INTERLIVE Network
• Systematic review/meta-analysis (2024) on wearable VO2max estimation
• 2025 validation study: Apple Watch VO2max vs indirect calorimetry
• Validation study: Garmin fēnix 6 VO2max estimates vs laboratory measurement (2025)

FAQ

What is Ronaldo’s VO2 max?

Specific individual figures like this aren’t publicly documented in verified research sources, so we can’t confirm a reliable number here. What’s well established is that elite endurance and team-sport athletes typically post high VO2max values, though wearable estimates for very fit individuals tend to carry more error than for recreational exercisers, as covered above.

What is the best wearable to track VO2max?

No single device has been shown to be consistently more accurate than the others across independent validation studies. Devices using exercise-based algorithms, drawing on heart rate and pace data from a run rather than resting data alone, tend to track closer to laboratory values, with the Garmin fēnix 6 showing a MAPE near 7.05% under controlled test conditions in one validation study.

How accurate is WHOOP’s VO2max estimate?

Independent, peer-reviewed validation data specifically on WHOOP’s VO2max estimate isn’t covered in the research summarised here, so a precise accuracy figure can’t be stated. As with other wearables, WHOOP’s estimate is best treated as a trend indicator rather than a clinical measurement, following the same exercise-based versus resting-based accuracy patterns seen across devices.

Can Fitbit calculate VO2max?

Yes, Fitbit estimates VO2max using a combination of heart rate data and, on models with GPS, pace during outdoor runs. Like other consumer wearables, this is an estimate rather than a direct measurement, so it’s best read as a trend over time rather than a one-off, precise figure.

How can I improve the accuracy of my wearable’s VO2max reading?

Testing consistently on the same flat route, warming up properly, and using a chest strap where possible all reduce the noise in your reading, as detailed in the protocol above. Entering a real, tested maximum heart rate rather than relying on the age-predicted default also helps, since a 15 beat-per-minute error in HRmax can shift the VO2max estimate by 7 to 9%.

Recommended

• Resting heart rate is the cheapest fitness test you own
• Your recovery score, decoded signal by signal
• The stairs test: 60 seconds that predicts a lot