The aim of this study was to analyse the evolution of motor performance in a group of patients who underwent surgical knee replacement. We also assessed patient satisfaction at 3 and 9 months after the operation. Sixty-two patients (40 women, 22 men, mean age 72.4 years) underwent isokinetic evaluation associated with surface EMG. All the patients had undergone a total prosthetic knee joint replacement for arthrosis, and had followed a standardised 30-day rehabilitation program at our centre for recovery of knee joint function, strengthening of the flexor-extensor muscles and restoration of ambulation as best as possible. Each of these patients underwent isokinetic evaluations 3 months and 9-10 months after their operation. The isokinetic test consisted of carrying out 5 flexion-extensions of the knee at an angular velocity of 60 degrees/second, followed by an endurance test of 30 repetitions of flexion-extension of the knee at 120 degrees/second. This isokinetic test provided data on extensor strength and flexor strength; furthermore, specifically designed software allowed simultaneous visualisation of the surface EMG tracing, torque and joint excursion. The first analysis showed a macroscopic decrease in the strength of the flexor-extensor muscles of the knee. This muscle weakness was clearly evident 3 months after the operation and was particularly pronounced for the extensor muscles of the operated limb compared with the muscles of contralateral, unoperated limb. Muscle weakness was still present 9 months after the operation although there had been a considerable improvement compared with 6 months previously; the imbalance in the flexor/extensor ratio, which differed from that in the contralateral, unoperated limb, was noted to be still present. The surface EMG demonstrated persistent myoelectrical activity in the flexors even when the extensor activity was predominant: this is an expression of imbalance between agonists-antagonists. A questionnaire administered to the patients about their satisfaction with the operation revealed that 9 months after the surgery 23 patients still complained of continuous joint pain with increased loads, 27 complained of frequent pain with load-bearing and only 12 no longer complained of any disturbance, manifesting full satisfaction with the operation. Total knee replacement is thus a valid treatment in those cases in which degenerative joint disease necessitates a radical solution. Nevertheless, dissatisfaction with the operation is common, and may be due to persistence of joint pain and incomplete joint recovery of joint function or muscle strength. In our series of patients, we found that although there was progressive recovery of strength of both the flexor and extensor muscles, a considerable imbalance remained between the operated and unoperated limbs; this finding was also confirmed by the surface EMG investigation.