This study investigated the contribution of muscle architecture to the differences in the torque-velocity and power-velocity relationships between older (OM n = 9, aged 69-82 years) and younger men (YM n = 15, aged 19-35 years). Plantarflexors' (PF) maximal isometric and concentric torques were recorded at 0.87, 1.75, 2.62, 3.49 and 4.36 rad s(-1). Physiological cross-sectional area (PCSA) was calculated as the ratio of muscle volume (determined by magnetic resonance imaging) to muscle fascicle length (Lf, measured by ultrasonography). GM PCSA and Lf of the OM were, respectively, 14.3% (P < 0.05) and 19.3% (P < 0.05) smaller than of the YM. In the OM, GM maximum isometric torque and maximum contraction velocity (Vmax), estimated from Hill's equation were, respectively, 48.5 and 38.2% lower (P < 0.001) than in the YM. At all contraction velocities, the OM produced less torque than the YM (46.3% of YM at 0.87 rad s(-1) to 14.7% at 4.36 rad s(-1), P < 0.001). Peak power (PP) of the OM was 80% lower than that of the YM and normalisation of PP to muscle volume only reduced this difference by 10%. Normalisation of torque to PCSA reduced, but did not eliminate, differences in torque between YM and OM (9.6%) and differences in torque/PCSA increased with contraction velocity (P < 0.05). After normalisation of velocity to Lf, the difference in Vmax between the OM and the YM was reduced to 15.9%. Thus, although muscle architecture contributes significantly to the differences in the torque- and power-velocity properties of OM and YM, other contractile factors, intrinsic to the muscle, seem to play a role. It is noteworthy that the deficit in PP between OM and YM is far greater than that of muscle torque, even after normalisation of PP to muscle volume. This finding likely plays an important role in the loss of mobility in old age.