Two hundred and forty-seven consecutive total knee replacements using the Insall-Burstein standard total condylar knee system with 2 years' follow-up were studied. Of the 238 available to follow-up, two were reoperated during the period, owing to infection and traumatic loosening of a tibia component respectively. Seventy-nine patients were operated for rheumatoid arthritis and 94 for osteoarthritis. In assessment of the total therapeutic result, registration of pain, walking ability, range of motion, muscle strength, flexion deformity, valgus-varus deformities, instability, and use of walking aids were included, according to a rating system. Excellent or good results were found in 91.5%; 6.5% were fair, and 2% were poor. There were two cases of infection; one resulted in removal of the prosthesis and an arthrodesis was done. There were four suspected aseptic loosenings, three cases of transient peroneal nerve palsies, and three patients who developed reflex dystrophy.