Considerable variability exists between people in their health-and performance-related adaptations to conventional endurance training. We hypothesized that some of this variability might be due to differences in the training stimulus received by the working muscles. In 71 young sedentary women we observed large variations in the ratio of one-leg cycling muscle aerobic capacity ((V) over dot(O2peak)) to two-leg cycling whole-body maximal oxygen uptake ((V) over dot(O2max); Ratio1:2; range 0.58-0.96). The variability in Ratio1: 2 was primarily due to differences between people in one-leg (V) over dot(O2peak) (r = 0.71, P <0.0005) and was not related to two-leg(V) over dot(O2max) (r = 0.15, P = 0.209). Magnetic resonance imaging (n = 30) and muscle biopsy sampling (n = 20) revealed that one-leg (V) over dot(O2peak) was mainly determined by muscle volume (r = 0.73, P <0.0005) rather than muscle fibre type or oxidative capacity. A high one-leg (V) over dot(O2peak) was associated with favourable lipoprotein profiles (P = 0.033, n = 24) but this was not the case for two-leg (V) over dot(O2max). Calculations based on these data suggest that conventional two-leg exercise at 70% (V) over dot(O2max) requires subjects with the lowest Ratio1: 2 to work their legs at 60% of single-leg (V) over dot(O2peak), whilst those with the highest Ratio1: 2 work their legs at only 36% of maximum. It was concluded that endurance training carried out according to current guidelines will result in highly variable training stimuli for the leg muscles and variable magnitudes of adaptation. These conclusions have implications for the prescription of exercise to improve health and for investigations into the genetic basis of muscle adaptations.