Measurements were made from annual follow-up radiographs, obtained over 27.6 years, of 860 cemented total hip arthroplasties implanted by one surgeon. Femoral components were made of stainless steel or titanium alloy, were non-modular, and were all fixed with cement, and acetabular cups were all-polyethylene and were fixed with cement. Radiographic outcome was correlated with the shape and material of the femoral component. Specifically, throughout the follow-up, stems made of titanium alloy were at greater risk of developing bone-cement radiolucent lines than those made of stainless steel, the difference ranging from approximately 10-50 percent at 2-10 years of follow-up. Similarly, titanium alloy stems were at greater risk of developing endosteal scalloping, indicating osteolytic lesions. Among the stainless steel Charnley cobra and straight-narrow Charnley stems, none developed cement fracture, only one became radiographically loose and one developed endosteal scalloping. The differences in the risk of developing radiolucent lines, cement fracture and progressive loosening among these stems were correlated with the relative rigidity of the femoral stems, and were generally consistent with the predictions made heretofore using finite element models, although differences in stem surface finish and femoral ball size and material could have also influenced the results.