Background: Division by total body weight is the usual way to standardise peak oxygen uptake (peak VO2) for body size. However, this method systematically underestimates cardiopulmonary fitness in obese individuals. Our aim was to analyse whether lean-mass is a better base for a body mass-independent standard of cardiopulmonary fitness. Methods: A population based sample of 578 men (body mass index (BMI) 19-47kg/m(2)) and 592 women (BMI 16-49kg/m(2)) 57-78 years of age. Peak VO2 was assessed by respiratory gas analysis during a maximal exercise test on a cycle ergometer. We studied the validity of the weight-ratio and the lean mass-ratio standards in a linear regression model. Results: The weight-ratio standard implies an increase of peak VO2 per additional kg body weight with 20.7ml/min (95% confidence interval (CI): 20.3-21.1) in women and 26.9ml/min (95% CI: 26.4-27.5) in men. The observed increase per kg is only 8.5ml/min (95% CI: 6.5-10.5) in men and 10.4ml/min (95% CI: 7.5-13.4) in women. For the lean mass-ratio standard expected and observed increases in peak VO2 per kg lean mass were 32.3 (95% CI: 31.8-32.9) and 34.6 (95% CI: 30.0-39.1) ml/min for women and 36.2 (95% CI: 35.6-36.8) and 37.3 (95% CI: 32.1-42.4) ml/min in men. The lean mass-ratio standard is a body mass-independent measure of cardiopulmonary fitness in 100% of women and 58% of men; corresponding values for the weight-ratio standard were 11% and 16%. Conclusions: For comparisons of cardiopulmonary fitness across different categories of body mass, the lean mass-ratio standard should be used.