To investigate the relationships between the age-associated decline in peak oxygen uptake (peak VO2), leg muscle mass (LMM) and leg strength (LS) under the modulating effect of gender and level of physical activity (LPA, as assessed by questionnaire), we evaluated 71 sedentary subjects [males(M):females(F), 34:37], aged 20-80 years. Peak VO2 at maximum cycle ergometry was related to LMM (dual energy X-ray absorptiometry) and LS (isokinetic dynamometry) using both standard (y x x(-1)) and power function ratios (allometry). Absolute values of all variables were higher in males and declined with age (P < 0.01). Differences between the genders disappeared after peak VO2 x LS(-1.46(M)or -1.62(F)) adjustment but remained significant between peak VO2 LMM-(-0.51 (M)or -0.45(F)) and LS LMM- (-091(M) or -1.10(F)) corrected values. Allometric correction of peak VO2 and LS by LMM slowed their age-related declines; the flattening effect however, was more evident in the females. Interestingly, while no age-related decrement in peak VO2 LS(-1) was found, power function ratio values also declined with age, and at a slower rate in females. These findings are consistent with those obtained following a multiple regression analysis using LPA as an independent covariate (P < 0.01). We conclude that using adequate scaling methods and controlling by LPA: (1) no gender differences are observed in peak VO2 adjusted for LS, (2) loss of LS and LMM are important determinants of the age-associated reduction in physical fitness in both genders but (3) age per se has an LS- and LMM-independent influence on the functional decline, although this intrinsic effect of senescence is less pronounced in females.