A previous study (Heil et al., MSSE 26(5):S63, 1994) found that submaximal oxygen uptake (˙VO2) was minimized as a quadratic function of preferred hip angle (HA) across nine combinations of seat-tube (70°, 80°, 90°) and trunk angle (10°, 20°, 30°) in competitive cyclists during submaximal cycle ergometry. These observations may have been the result of training-specific alterations in the hip and/or knee extensor torque capabilities. The present study evaluated this theory by creating theoretical torque-angle curves for individual muscles participating in either hip or knee extension during the power-phase of bicycle pedaling. Sagittal-view kinematics of the hip and knee joints from the previous study(n=13) were used as input to the model. Based upon a two-component Hill-type model, the maximal theoretical torques (TMAX) of four hip (gluteus maximus (GM), 3 hamstrings) and four knee extensors (rectus femoris (RF), 3 vasti) as a function of changes in joint and crank angles were calculated. The algorithms for TMAX combined muscle, tendon, and sarcomere architecture data from the literature with moment-arm length and force-length algorithms. The results indicated: 1) Changes in TMAX were greatest for GM and RF; 2) TMAX magnitude for GM and RF increased with an increase in mean HA, while peak TMAX values occurred earlier in the crank cycle for GM and later for RF. Thus, TMAX magnitude and timing for GM and RF may explain the previously observed relationship between preferred HA and˙VO2.