Of the first 300 consecutive patients who had a Charnley total hip replacement at the Mayo Clinic during the years 1960 to 1970, 207 (231 hips) were re-evaluated ten years postoperative by questionnaire and roentgenograms. Forty-three of these were also evaluated by personal examination. Roentgenographic loosening of the acetabular component was determined using the criterion of a complete radiolucent line more than one millimeter in width at the bone-cement interface or any migration or tilting of the component. For the femoral component, the criterion for loosening was a radiolucent line more than one millimeter wide at either the bone-cement or the cement-prosthesis interface, or any change in the position of the component. As previously reported, the incidence of loose components at five years was 6.5 per cent for the acetabular component and 24 per cent for the femoral component. At ten years the incidence of loosening had increased to 11.3 per cent for the acetabular component and 29.9 per cent for the femoral component. Therefore, between five and ten years postoperatively the rate of femoral loosening decreased, while the rate of acetabular loosening remained about the same. The overall-revision rate for loosening of total hip components increased from 3 per cent at five years to 7.4 per cent at ten years. Acetabular wear was not a significant problem. Resorption of the medial femoral cortex near the calcar was generally non-progressive and was not significantly related to loosening. two modes of loosening are suggested, the more common being cracking of the cement mantle due to circumferential (hoop) stresses within the cement. This series probably represents a so-called worst-case experience, since changes in design and materials as well as the improvements in surgical technique that have evolved over the past decade should provide significantly better long-term fixation.