Today the following two methods of measuring the bone mineral density (BMD) are widely used: dual-photon absorptiometry with X-rays (DEXA) and quantitative computed tomography (qCT). DEXA measures the BMD of the vertebral column, the femur and the whole skeleton, qCT the vertebral column and the radius and tibia. The qCT method permits a distinction to be made between trabecular and cortical bone, which offers an advantage in terms of the choice of treatment because cortical and trabecular bone do not change in parallel. For example, fluorides are contraindicated when cortical bone is diminished. There are important differences among the methods of densitometry concerning reproducibility. The highly sensitive method of peripheral qCT (Densiscan) has an overall reproducibility of 0.3 % (mixed population), allowing a change in BMD of less than 1 % to be detected. This enables us to adapt quickly the treatment according to the stage and progression of the disease. In the perimenopause patient, bone mineral densitometry is indicated to detect high-risk patients; in the postmenopause patient, densitometry should be performed in order to ''tailor'' the treatment of osteoporosis. Patients with a ''slow loser'' osteoporosis are preferably treated with bone-stimulating substances (e. g., fluorides, ossein-hydroxyapatite compound), patients with a ''fast loser'' osteoporosis (trabecular bone loss, measured at the distal radius, of more than 3.5% per year) with antiresorptive agents such as estrogens, bisphosphonates, calcitonin and anabolic steroids. This tailored treatment reduces not only the number of nonresponders, but improves patient compliance because it allows the choice, duration or change of the treatment to be substantiated by means of hard data. Beside the quantitative parameters of bone densitometry, qualitative examination of the bone is of great importance since the stability of bone depends on its structures. The latter can be visualized and determined by qCT due to the spatial resolution of 0.2 mm of the Densiscan.