Fifty-two Stanmore hinged knee replacements were used as revision arthroplasties for failed arthroplasties. The patients were followed for a mean of 44.7 months. On clinical examination, 23% had good, 48% fair, and 29% poor results. Worst results were achieved when revising bulkier or more constrained primary knee joints. Two patients exhibited radiologic evidence of loosening, but at the latest follow-up examination no patients had clinical evidence of loosening. There was an infection rate of 4%. In view of the poor results and the advent of a new generation of knee arthroplasty devices, hinged knee implants may have only a limited place in revision arthroplasty.