WHOOP VO2 Max Accuracy: Why Yours Looks Wrong

Last updated: August 20, 2026

WHOOP is the only company that has tested WHOOP's VO2 max. Its own validation against a metabolic cart in 248 people puts the estimate within 3.3 to 3.7 mL/kg/min of the lab value, which would be genuinely good if it holds. Two things stop that being the whole answer. Nobody outside WHOOP has ever checked it, and WHOOP itself reports that the error grows at the top and bottom of the fitness range, which is exactly where the people asking this question tend to sit. If your number looks wrong, though, the cause is usually more boring than the accuracy question: a device that cannot produce it, a gate you have not passed, or a model you quietly dropped out of.

Want the cross-device version of this question? The VO2 max accuracy tool lets you filter the validation studies by brand and method, then come back for what is actually going on with your WHOOP.

I started digging into this because the pages ranking for it either quote WHOOP's accuracy figure without the caveat printed two paragraphs below it, or skip the mechanics entirely and never explain why half the people asking have no number at all.

What WHOOP's own testing actually found

WHOOP ran a graded treadmill test to volitional exhaustion against a Parvomedics TrueOne 2400 metabolic cart, n=248, gas exchange sampled every 30 seconds. Maximal effort was confirmed on four criteria: a VO2 plateau, RPE above 17, heart rate within 10% of age-predicted max, and RER at or above 1.10. Tests that failed those criteria were thrown out of training.

Model	Mean absolute error	MAPE	Pearson r
Passive (default)	3.7 mL/kg/min	8.0%	0.90
GPS-augmented	3.3 mL/kg/min	7.1%	0.91

Credit where it is due: that is a properly run protocol, and WHOOP says it used the INTERLIVE consortium's 2022 recommendations as the blueprint, which is the methodological review that flagged inconsistent maximal-effort confirmation as a persistent gap in this field. WHOOP built to the standard the field asked for.

Two caveats change how you should read those numbers.

The first is WHOOP's own, in its own words: "The passive model showed slightly higher error in the lowest and highest true-VO2 Max brackets." Read the surrounding claim carefully, because it is narrower than it first appears. Mean absolute error stayed under 4 mL/kg/min across sex, across age bands and across true-VO2 bands in the GPS-augmented model. The extremes problem belongs to the passive model, which is the default and the one most members are on.

The second is that all of this is internal and cross-validated, not peer reviewed. Checking Europe PMC for WHOOP validated against maximal oxygen uptake returns five records, none of which tests WHOOP's VO2 max estimate against a lab CPET. Independent peer-reviewed VO2 max validation currently exists for four brands only: Garmin, Apple, Polar and Fitbit. WHOOP is not among them, and neither are Samsung, Oura, Coros or Suunto.

So the honest headline: WHOOP's own testing says within 3 to 4 mL/kg/min. Nobody has independently confirmed it. And WHOOP says the error is worst at the extremes.

Why you might not have a VO2 max at all

This is the most common version of "it looks wrong," and it has nothing to do with accuracy.

VO2 Max only exists on WHOOP 5.0 and MG. A WHOOP 4.0 will never show it, at any membership tier, ever. WHOOP 5.0 and MG launched on 8 May 2025. If you are on older hardware, that is the answer and there is no setting to change.

If you do have the right device, every tier includes it. Healthspan and WHOOP Age start at Peak, but VO2 Max does not.

Tier	Price	Device	VO2 Max included
One	$199/yr	WHOOP 5.0	Yes
Peak	$239/yr	WHOOP 5.0	Yes
Life	$359/yr	WHOOP MG	Yes
Any tier on WHOOP 4.0	n/a	WHOOP 4.0	No, and no upgrade path in software

Then there is the unlock gate, and it catches people out because it is about sleep rather than exercise. You need roughly three weeks of consistent overnight wear. WHOOP words this two ways across two pages: its support article says at least 14 sleeps in 21 days, its accuracy page says at least 14 recoveries in the previous 21 days. A recovery requires a scored sleep, so these are near-equivalent, but they are not identical wording and it is worth knowing both exist. Once you pass the gate, WHOOP backfills up to 6 months of estimates depending on how much data you have accrued, which is why the number can appear and then immediately show a history you never watched build.

The silent 90-day downgrade nobody mentions

This is the finding that explains a lot of "my number got weird over winter" complaints, and it is buried on WHOOP's accuracy page rather than its support article.

A GPS-tracked outdoor run of 15 minutes or more improves your estimate but is not required to have one. That much is well documented. What is not well documented: the GPS-augmented model only applies if you logged an outdoor GPS run in the last 90 days.

So the better model expires. Stop running outdoors for three months, whether that is weather, injury, or a winter spent on the treadmill, and you silently fall back to the passive model. Which is the model with the extremes problem. Nothing in the app announces this.

Two related limits worth knowing while you are there. Elevation and terrain do not yet affect the estimate, so a hilly or trail run gets no credit for being hard, and a trail runner's number can read low for a purely structural reason. And WHOOP does not read Strava or Apple Health for VO2 max calculations, so an external GPS watch will not feed it no matter how well that watch tracks. If you want the GPS model, the run has to be recorded on WHOOP, outdoors, with "track route" toggled on after you tap Start Activity, on flat ground.

If none of that is available to you, the metrics WHOOP does produce every single day still respond to what you eat and drink. Kygo pulls your WHOOP recovery, HRV, resting heart rate and sleep next to your food log and finds which patterns are actually yours, free on iOS and Android.

Every reason yours looks wrong, and what to do

What you are seeing	Why	What to do
No VO2 max at all	Device is WHOOP 4.0 or older	Only 5.0 and MG produce it. No fix
No VO2 max at all	Fewer than 14 sleeps logged in 21 days	Wear it overnight. The gate is sleep, not workouts
It has not changed in days	It updates weekly, usually Tuesday	Not a bug. WHOOP calls it a very slow moving metric. Judge it monthly
It jumped the moment you got it	Backfill of up to 6 months of accrued data	Expected behaviour
Reads low if you are very fit	WHOOP reports higher error in the highest true-VO2 bracket	No user-side fix. Enter a lab value manually to calibrate
Reads off if you are very unfit	Same admission, lowest bracket	Same
Wrong after a weight change	Weight is a direct model input	Update weight in the VO2 Max trend view, and check the units
Wrong if you are a masters athlete or a teenager	Max HR is an input, and 220 minus age fits individuals badly	Set your max HR manually
Got vaguer over winter	No outdoor GPS run in 90 days, so you dropped to the passive model	Log one flat outdoor GPS run of 15 minutes or more
Trail and hill running not reflected	Elevation and terrain are explicitly not used yet	Log a flat GPS run to feed the better model
Disagrees with Garmin or Strava	No integration, and every brand uses a different algorithm	Do not cross-compare brands
Under 18, or pregnant	WHOOP states it is validated for 18 and over, and not validated during pregnancy	Treat as directional only

Two things it is worth clearing up while you are here. VO2 Max is not an input to your Recovery score, so a low VO2 max is not dragging your recovery down; if you want what actually feeds Recovery we broke it down in WHOOP stress score and recovery explained. And manual entry is supported, but WHOOP recommends entering only lab-assessed values, which then calibrate the trend rather than overwrite it.

The mechanism underneath most of the accuracy complaints is worth stating plainly. A VO2 max estimate leans on heart rate during hard effort, which is exactly where wrist optical sensors are weakest, and on an assumed maximum heart rate, which INTERLIVE's 2022 review flags as a leading error source. WHOOP is wrist or bicep optical with no chest-strap input path and no external GPS import, so the two mitigations that exist elsewhere in this category are both unavailable to it. That is an observation about inputs, not a measured error figure for WHOOP.

The very fit and the very unfit get the worst readings

WHOOP's admission about the extremes lines up with something the independent literature has found on other devices, and the way to say this carefully matters.

Study	Sample	Finding
Engel 2026, Garmin Forerunner 245	35 endurance athletes, split by training status inside one sample	Highly trained MAPE 9.4 to 10.4% against moderately trained 2.8 to 4.1%. Error 2.5 to 3.4 times higher
Carrier 2023, Garmin Fenix 6 plus HRM-Run chest strap	20 participants, all at or above the 95th percentile for aerobic capacity	MAPE 6.85%. No degradation

The fair summary is that accuracy degrades in highly trained users, and the one elite study that avoided this used a chest strap. That is a hypothesis about why, not a proven mechanism: nobody has run wrist against strap in the same elite sample.

Two things this is not. It is not "all wearables get worse the fitter you are," and those percentages are not WHOOP's. They come from one wrist-based Garmin device in one study. What makes them worth mentioning at all is the convergence: independent research on other wrists points at degradation at the top of the range, and WHOOP's own validation independently reports higher error in exactly the same bracket. Two different sources, same edge of the distribution.

There is also no independent evidence that WHOOP reads high or low as a direction. WHOOP's own data reports absolute error, not signed bias by group, so anyone telling you WHOOP systematically under-predicts trained users is going beyond what has been measured.

Why it disagrees with your Garmin or Apple Watch

Because there is no shared data and no shared method. WHOOP does not read Strava or Apple Health for VO2 max, so the two devices are estimating independently from different inputs with different proprietary models. A gap between them is the expected result, not evidence that either is broken.

Method class explains more than brand does. Exercise-based modelling, where the device watches your heart rate response to a known pace, consistently beats resting or non-exercise estimation. That is visible inside WHOOP's own numbers: the GPS-augmented model at 7.1% against the passive model at 8.0%. Across brands, Garmin sits around 7% error against lab CPET in general populations and is the most replicated. For the full ranking and what each study actually tested, see most accurate VO2 max wearable.

Worth noting the same troubleshooting logic applies elsewhere in the category for structural rather than accuracy reasons. The Fitbit Air, for instance, often produces no VO2 max at all because it has no GPS chip, which we covered in why Fitbit Air VO2 max says not tracked.

The practical rule: compare each device only against itself, over months.

Common questions

How accurate is WHOOP VO2 max?

WHOOP's own testing against a metabolic cart in 248 people puts it within 3.3 to 3.7 mL/kg/min of the lab value, with MAPE of 7.1% using GPS and 8.0% passive, and r around 0.90. That testing is internal and has not been peer reviewed, and no independent study has validated it.

Is WHOOP VO2 max accurate for athletes?

WHOOP reports higher error at the highest true-VO2 bracket. Independent research on other wrist devices points the same way, and the one elite-athlete study that showed no degradation used a chest strap, which WHOOP has no equivalent for.

Why is my WHOOP VO2 max so low?

Usually an input problem rather than a measurement problem: outdated weight, a wrong max HR, or an estimate built only from the passive model. Log a flat 15-minute outdoor GPS run with track route on, update your weight and max HR, and add a lab value if you have one.

Why don't I have a VO2 max on WHOOP?

VO2 Max is only available on WHOOP 5.0 and MG, and it needs at least 14 sleeps logged in 21 days before it will appear.

How often does WHOOP update VO2 max?

Weekly, usually on Tuesdays. WHOOP describes it as a very slow moving metric, so a reading that has not moved since last week is working correctly.

Does WHOOP need a GPS run for VO2 max?

No. An outdoor GPS run of 15 minutes or more is not required to get an estimate, but it unlocks WHOOP's more precise GPS-augmented model, and that model only applies if you logged such a run within the last 90 days.

Why is my WHOOP VO2 max different from my Garmin or Apple Watch?

Different algorithms, different inputs, and no shared data. WHOOP does not read Strava or Apple Health for VO2 max. Compare each device only against itself over time.

Does VO2 max affect my WHOOP Recovery score?

No. WHOOP states that VO2 Max is not an input to Recovery.

Does WHOOP count hills or trail runs?

Not yet. WHOOP states that elevation and terrain do not currently affect the estimate, so a hard hilly run is not credited as harder than a flat one.

The bottom line

WHOOP's VO2 max is probably fine for most people, on WHOOP's own numbers, and those numbers come from a well-run protocol. The problem is that they are the only numbers, and WHOOP has told you where they are weakest. If you are extremely fit, extremely unfit, under 18, pregnant, or running on hills, the estimate is doing something harder than it was validated for.

Before you conclude the number is broken, work through the boring causes in order: right device, 14 sleeps logged, weight and max HR current, and an outdoor GPS run inside the last 90 days. Most complaints resolve at one of those four. After that, treat it as a trend line you read monthly, not a score you check on Tuesdays.

If you want to actually move the number rather than audit it, the levers are the same whichever device reports it, and we ranked them by evidence in what affects VO2 max. Kygo sits alongside WHOOP and connects the metrics that do move day to day, recovery, HRV, resting heart rate and sleep, to what you actually ate. Free on iOS and Android.

---

Key sources: WHOOP's VO2 Max support article for the device restriction, the 14-sleeps gate, the weekly update cadence, backfill, the input list, the GPS rule, elevation and terrain, the under-18 and pregnancy limits, and the Recovery clarification; WHOOP's "How Accurate Is WHOOP VO2 Max" page for the n=248 validation, the passive and GPS-augmented figures, the 90-day GPS window and the extremes admission, page dated 22 April 2026; WHOOP's membership page for tier pricing; Engel et al. 2026 (European Journal of Applied Physiology 126:591-603, PMID 40770433) and Carrier et al. 2023 (Technologies 11:71) for the trained-athlete comparison; Molina-Garcia et al. 2022 (the INTERLIVE meta-analysis) for assumed max HR as a leading error source and for the exercise-based versus resting-based method split; and a Europe PMC search run 19 August 2026 confirming no independent WHOOP VO2 max validation exists.

---

A note on what we left out. We did not publish the widely repeated claim that WHOOP under-predicts trained runners, or the mean absolute error range attached to it, because both trace to user field reports rather than any study. We did not use the 5% figure often quoted for Garmin, which is a vendor claim; the independent Garmin figures are 6.7%, 6.85% and 7.05%. We did not repeat the Samsung Galaxy Watch VO2 max accuracy figure that circulates widely, because it corresponds to no indexed study. And we did not carry over WHOOP's mortality and survival claims from its accuracy page, because we have not checked them against the primary sources.

Written for general information, not medical or training advice. VO2 max estimates from any consumer wearable are approximations.

If you have a lab CPET result and a WHOOP estimate from the same month, I would like to know how far apart they were, because that comparison is the one nobody outside WHOOP has published.