Femoral bone remodeling after total hip replacement was studied by following patients who received 326 Charnley femoral prostheses for 10 to 20 years (mean, 13.3 years). The radiographic state of the bone remodeling was visually assessed and measured with a digitizer. Demineralization that started proximally and then progressed distally caused cortical thinning, which correlated with widening of the intramedullary canal, not with changes that developed in the periosteal width, and occurred in the medial femoral neck, around the proximal half of the stem, and around the distal half in 87%, 33%, and 10%, respectively. Cortical thinning around the distal half of the stem was always accompanied by proximal thinning, and extensive cortical thinning (both proximal and distal) correlated with both lower clinical scores and radiologic loosening of the femoral prosthesis. A low canal flare index of Noble, a large canal width, and a patient age of 60 years or more were risk factors for extensive cortical thinning. Accelerated polyethylene wear was related to resorption of the medial femoral neck but not to cortical thinning or radiological loosening. Cortical thickening occurred only around the distal half of the stem in 29%. These findings establish a basis for the performance of cemented femoral prostheses, and allow comparison of bone remodeling when evaluating other femoral prostheses.