SYNTH-78 · Guide

Estimating VO2max from cycling power

Give a range with its error (at least ±10 mL/kg/min), never a single number, and steer training by the power itself

Blimited evidenceMostly from trained to elite athletesWritten before October 2026, being re-checked

Decision

Give a range, not a number: VO2max ≈ 16.6 + 8.87 × best 5-minute W/kg, quoted with at least ±10 mL/kg/min and the note that the equation read about 7 too low when checked independently; set training from the power, and send anyone who needs the real value to a laboratory test.
In detail:

1. From best 5-minute power: VO2max ≈ 16.6 + 8.87 × W/kg REF-1015. Quote it as "roughly", with a band of at least ±10 mL/kg/min, and add that in an independent sample the same equation read about 7 too low REF-1016.

2. From ramp-test peak power the best equation still carries about ±12 mL/kg/min for an individual trained cyclist REF-1017.

3. Do not compare two athletes by the estimate, do not call a change of a few mL/kg/min progress, and do not set zones from it — zones come from the power.

4. If the athlete needs the real number (a health question, team selection, curiosity that a range does not satisfy), the answer is a laboratory test.

The progress marker is the same effort repeated: best 5-minute power, or power at a fixed heart rate, against the previous block.

Evidence

Equation and its fit in 46 road cyclists: R² 0.61–0.77 from a 5-minute test REF-1015.
External validation in 49 cyclists: bias −6.6 mL/kg/min, limits of agreement −17.2 to +3.8 REF-1016.
580 competitive cyclists: r = 0.68 in men between power-based prediction and measured VO2max, limits of agreement −15.5 to +8.3 REF-1017.

Practical Application

A good answer has three parts: the 5-minute power in W/kg, the range it implies, and one sentence on why the range is wide — efficiency and the anaerobic share of a 5-minute effort differ between riders.
The number on a watch or head unit is an estimate of the same kind and deserves the same caution.

Limitations

All three studies are in cyclists; no equivalent validated estimate from running pace is covered here.
The validated equation failed its only external test; no better field equation was verified.