Metabolic disorders often cause changes in the mineral content of the skeleton or the fat content of the bone marrow, or both. Non- invasive x-ray techniques have been developed for analyzing the mineral content of the skeleton and are frequently employed in the diagnosis of diseases such as osteoporosis. A technique which could in addition analyze fat content could improve the quality of the diagnosis and have wider application. Recently, we reported the development of two quantitative CT (QCT) methods that analyze both bone mineral and fat content. The present study is designed to further develop and apply these methods. Three major investigations are proposed. First, experiments will be performed to optimize the QCT methods with respect to accuracy, precision and x-ray dose. These experiments will be facilitated by using anthropomorphic lumbar sectional phantoms with plastic inserts that mimic vertebrae. Next, the accuracy of the techniques will be tested in excised vertebrae from human cadavers. The mineral estimates will be compared with those of other CT and photon absorptiometry techniques. The fat estimates will be compared with those of two nuclear magnetic resonance techniques. The actual mineral content of the vertebral sections will be determined by ashing and the fat content by fat extraction. Finally, the methods will be applied in selected human subject studies. Among these will be investigations of: the short-term reproducibility of the methods, the in vivo/in vitro accuracy of the methods when applied to the femoral head in hip replacement patients, and the correlation of the methods with others in evaluating normals and osteoporotics.