Fifty patients requiring bilateral total hip arthroplasty underwent a concomitant conventional hip arthroplasty on one side and an articular resurfacing procedure on the other, done by the senior one of us (M. A. R.). The average age of the patients was sixty-two years (range, twenty-one to eighty-seven years), and forty-seven of them were followed with serial radiographs and clinical evaluation for one year or more. The length of follow-up was five years for thirty-four patients with both prostheses intact. The average pain score for all hips at three years postoperatively was 5.5 points. Although the majority of patients at each follow-up interval did not prefer one procedure to the other, the conventional arthroplasty was significantly superior for those who had a preference, and radiographic evaluation revealed a statistically significant increased incidence of acetabular lucency at the bone-cement interface of the resurfacing arthroplasty at one, two, three, five, and seven years postoperatively (p less than 0.002). Two of the conventional prostheses were revised during this time-period (one because of infection and one, a broken stem) whereas thirteen (26 per cent) of the resurfacing prostheses were revised (eight hips had femoral loosening, five with concomitant acetabular loosening; three had acetabular loosening; and one had a femoral neck fracture) at an average of fifty-two months postoperatively (p less than 0.001).