1100 total hip prostheses were inserted between 1977 and 1978. Among them, the authors have selected at random 150 cases operated on by the Austin Moore approach and 150 cases by trochanteric division. The pathology and the type of patient were the same in both groups. A study of the late results showed no difference between the two series. In each, there were two early infections and three late infections. The rate of post-operative dislocation was the same, two early and six late in the Moore approach group, and five early, and two late in the trochanteric division group. There were six sciatic involvements after the Moore approach but only three after trochanteric division, but in this last group there were complications relating to the trochanteric division itself. There were no statistical differences in a comparison of the numbers of re-interventions required. Both techniques appeared to be satisfactory. The Moore approach appeared to be better for the comfort of the patient and does not give any additional risk. Trochanteric osteotomy is indicated in particularly stiff hips, when the greater trochanter has to be displaced downwards or in cases of post-operative scarring.