Eighty-nine Charnley total hip arthroplasties were performed with cement, by a single surgeon, in sixty-seven patients who were less than fifty years old at the time of the procedure. The follow-up period ranged from sixteen to twenty-two years (average, eighteen years). The most recent evaluation consisted of a clinical and a radiographic examination for forty-six patients (fifty-eight hips), a telephone interview and a review of the most recent radiographs for eleven patients (sixteen hips), and a telephone interview only for six patients (ten hips). Four patients (five hips) were lost to follow-up. Eleven (13 per cent) of the eighty-four hips that were followed were revised because of aseptic loosening of the acetabular implant, and two hips (2 per cent) were revised because of aseptic failure (loosening or fracture) of the femoral component. When the number of hips that were revised was combined with the number of hips in which there was radiographic evidence of failure, the rate of loosening of the acetabular component was 50 per cent (forty-two hips) and the rate of failure of the femoral component was 8 per cent (seven hips). Three hips had recurrent dislocations after the operation; none of the three was revised. One patient (two hips) had Grade-III heterotopic ossification according to the system of Brooker et al.; no other patient had more than Grade-II heterotopic ossification. Nine hips had a trochanteric non-union without migration of the trochanter.(ABSTRACT TRUNCATED AT 250 WORDS)