We selected randomly a consecutive series of 162 patients requiring hip replacement to receive either a cementless, hemispherical, modular, titanium acetabular cup or a cemented, all-polyethylene cup. These replacements were performed by two surgeons in four general hospitals. The same surgical technique was used and a 26 mm metal-head femoral component was used in every case. After exclusions, 115 hips were studied for differences in rates of wear and osteolysis. The mean clinical follow-up was eight years and the mean radiological follow-up, 6.5 years. The cementless cups wore at a mean rate of 0.15 mm per year and the cemented cups at 0.07 mm per year. This difference was significant (p < 0.0001). Our findings in this mid-term study suggest that cementless cups wear more than cemented cups.