Conventional radiographs, technetium-99m bone scans, and gallium-67 scans were reviewed in 44 patients who had undergone cemented total hip joint replacement and were imaged because of suspicion of prosthesis loosening or infection. A complete radiolucent line of 2 mm or wider along the bone-cement interface or metal-cement lucency on conventional radiographs was used as the criterion for prosthetic loosening with or without infection and proved to be 54% sensitive and 96% specific. Scintigraphic criteria for prosthetic loosening were increased focal uptake of the radiopharmaceutical for the femoral component and increased focal or diffuse uptake for the acetabular component. For bone scintigraphy, sensitivity was 73% and specificity was 96%. Combining the results of conventional radiographs and bone scans increased sensitivity to 84% and decreased specificity to 92% for the diagnosis of loosening, infection, or both. The study also showed that Ga-67 scintigraphy has a low sensitivity for the detection of infection.