One hundred consecutive Muller curved-stem total hip replacements were reviewed ten years after operation. Twenty patients with twenty-two arthroplasties had died within the ten-year period without having a revision, and twenty-five arthroplasties had been revised for various reasons. Of the remaining fifty-three arthroplasties, thirty-five were classified as good or excellent, with Harris hip scores of 80 points or higher, and eighteen were classified as poor or fair, with scores lower than 80 points. Follow-up radiographs, made for all but six of the fifty-three hips at ten years, showed a 23 per cent incidence of migration of the acetabular component and a 28 per cent incidence of migration of the femoral component. In addition, there was a 15 per cent incidence of bone resorption in the proximal end of the femur without migration of the femoral component and a 4 per cent incidence of osteolytic defects about the femoral component, also without migration. Combining the radiographically loose replacement (migration) with the clinically loose ones (revised), the over-all incidence of aseptic loosening was 29 per cent for the acetabular component and 40 per cent for the femoral component. There was a positive correlation between the incidence of loosening of the femoral component and younger age, heavier weight, male sex, unilateral hip disease, a wide femoral canal, and varus position of the femoral component, whereas the incidence of loosening of the acetabular component was increased only in association with older age. The rate of loosening of the femoral component appeared to be higher during the early follow-up period and to decrease with time, while the rate of loosening of the acetabular component appeared to be lower during the early follow-up period but to increase with time.