Abstract Background Autologous chondrocyte implantation has become an established technique for addressing knee cartilage defects. Despite reported improvement in pain and regeneration of hyaline-like repair tissue, little has been reported on the recovery of knee strength. Methods Knee strength assessment was undertaken in 60 patients at 5 years following autologous chondrocyte implantation. Using an isokinetic dynamometer, and during isokinetic knee extension and flexion angular velocities of 60°, 90° and 120°/s, the peak torque, torque at 45° of knee flexion and hamstrings/quadriceps ratio was obtained, in both the operated and non-operated limbs. Pain at the time of assessment was obtained. Independent sample t-tests were used to assess differences in the operated and non-operated sides. Findings There were no significant differences (p > 0.05) between the operated and non-operated legs in the peak knee flexor torque or knee flexor torque at a knee flexion angle of 45°, at all angular velocities (60°, 90° and 120°/s). While the peak knee extensor torque was less in the operated leg at all angular velocities, these differences were not significant (p > 0.05). However, a significantly reduced (p < 0.05) knee extensor torque at a knee flexion angle of 45°, was observed at all speeds. Interpretation While patients had recovered their knee flexor strength, they still demonstrated a reduced knee extensor strength profile at 5 years. This demonstrates that the early supervised rehabilitation phase following autologous chondrocyte implantation is not sufficient to restore long-term knee strength, and ongoing patient advice and rehabilitation is required extending beyond this early period. It is unknown how this prolonged reduction in strength may affect long-term graft outcome.