VO2 max is quoted casually by fitness watches, but the measurement it refers to is a laboratory procedure. Understanding the real test explains why the estimated version moves around.
What the number represents
VO2 max is the maximum rate at which a body can take in oxygen and deliver it to working muscle, expressed per kilogram of body weight per minute.
It is a ceiling on aerobic energy production, because sustained work depends on oxygen reaching mitochondria fast enough to keep pace with demand.
Above that ceiling, effort continues briefly on anaerobic pathways that produce fatigue quickly, so the value marks where sustainable work ends.
The laboratory procedure
A subject exercises on a treadmill or bike while workload increases in stages, breathing through a mask connected to a gas analyser.
The analyser compares oxygen and carbon dioxide in inhaled and exhaled air, calculating consumption directly. The test continues to voluntary exhaustion.
A valid result requires the oxygen consumption curve to flatten despite rising workload, which is what confirms a true maximum rather than a stopping point.
How wearables estimate it
A watch has no gas analyser. It infers the value from the relationship between heart rate and pace or power during ordinary activity.
The logic is that a fitter person runs faster at any given heart rate, so the ratio between the two carries information about aerobic capacity.
Because the estimate depends on assumed maximum heart rate, on accurate pace and on running rather than walking, its error is substantial in absolute terms.
Trend is more useful than value
Errors in an estimate that come from stable assumptions tend to be consistent, so the direction of change is more trustworthy than the figure itself.
A device that shows a rising number over months is probably detecting a real improvement, even if the absolute value would not match a laboratory test.
Comparing figures between devices or between people is where the estimate stops carrying meaning.
The limits of the ceiling
Aerobic capacity is partly inherited, and the range of response to identical training is wide. Two people following the same programme will not converge.
Capacity also declines gradually across adulthood, though continued training slows the rate of decline considerably.
Performance depends on efficiency and sustainable fraction of capacity as much as on the ceiling, which is why the highest measured value does not always win a race.