Thirty-eight St. Georgs hinged-knee arthroplasties were studied with a follow-up of 2.5-8 years; 86.8% of the patients obtained an excellent or a good result, and 7.9% of the arthroplasties were failures. There was an infection rate of 3.8% and 3.8% loose prostheses. Among the failures there were two infected arthroplasties. One was arthrodesed, resulting in a pseudoarthrosis and one was successfully revised. One traumatic loosening lead to arthrodesis. It is concluded that hinged-knee arthroplasty under special circumstances may be a useful procedure.