We evaluated the results of 330 total hip arthroplasties that were performed with use of the Charnley prosthesis and cement in 262 patients by the senior one of us between July 1970 and April 1972. All hips had been thoroughly assessed preoperatively to document the patient's functional level. All patients had been disabled because of pain in the hip or a fracture of the hip, and 212 patients (81 per cent) had used walking aids. At a minimum of twenty years after the index operation, eighty-three patients (ninety-eight hips) were still living, 174 patients (224 hips) had died, and five patients (eight hips) had been lost to follow-up. The outcome of the arthroplasty was determined for all except the five latter patients. Thus, the outcome of 322 (98 per cent) of the 330 arthroplasties was known at the latest follow-up evaluation. Radiographs were available for sixty-three of the eighty-three patients (seventy-six [78 per cent] of the ninety-eight hips) who were alive for the entire follow-up period. Of the ninety-eight hips in the living patients, eighty-three (85 per cent) caused no pain, fourteen (14 per cent) caused mild pain, and one (1 per cent) caused moderate pain. Fifty-two hips (53 per cent) were in patients who did not use walking aids, and only seven (7 per cent) were in patients who used support for walking because of the hip. At the minimum twenty-year follow-up, thirty-two (10 per cent) of the 322 hips that had been followed had been revised: eight (2 per cent), because of loosening with infection; twenty-one (7 per cent), because of aseptic loosening; and three (1 per cent), because of dislocation. Of the ninety-eight hips of the patients who were still alive, fifteen (15 per cent) had been revised: three (3 per cent), because of loosening with infection; eleven (11 per cent), because of aseptic loosening; and one (1 per cent), because of dislocation. The rate of revision due to aseptic loosening of the acetabular component in all 322 hips was 6 per cent (eighteen hips), while in the ninety-eight hips of the patients who were alive at least twenty years after the arthroplasty, it was 10 per cent (ten hips). The rate of revision because of aseptic loosening of the femoral component in all 322 hips was 2 per cent (eight hips), while in the ninety-eight hips of the living patients, it was 3 per cent (three hips).(ABSTRACT TRUNCATED AT 400 WORDS)