Breast and prostate carcinomas are the tumors most commonly associated with skeletal metastases, and the skeleton is the most common site of metastatic disease and of first distant relapse in breast cancer. Bone metastases are the source of considerable morbidity, including pain and functional disability, fractures, hypercalcemia, and epidural compression. The classical radionuclide bone scan remains the most effective tool for the screening of metastatic bone disease, but X-rays are more specific and remain the essential tool for the diagnosis and characterization of bone metastases. Computed tomography is much more useful to diagnose early metastatic involvement of bone, particularly of the spine. Patients with exclusive skeletal metastatic involvement are still frequently excluded from classical therapeutic trials because of the difficulties in the assessment of response. Re-calcification of osteolytic lesions is indeed required when defining an objective response, but this criterion is insensitive and not quantitative. Moreover, the development of new osteoblastic lesions is often of difficult interpretation. A concomitant bone scan will help, but the absence of quantification of the changes and the “flare” phenomenon limit the usefulness of the technique. Pain and quality of life constitute simple, but frequently neglected, parameters of response to therapy. The clinical utility of tumor markers and of biochemical markers of bone turnover should also be more fully investigated. Neoplastic osteolysis is essentially mediated by the osteoclasts, which seem to be activated, maybe indirectly through the osteoblasts, by some tumor products. Various substances of tumoral origin have been proposed as mediators for this osteoclast activation, such as transforming growth factors, prostaglandins, and, more recently, products of the immune cells or parathyroid hormone-related peptide. Several current clinical attitudes for the treatment of metastatic bone disease are not based on properly conducted trials, such as the administration of postoperative radiotherapy or the fractionation of radiotherapy over two to four weeks. Bone metastases reportedly respond less well to endocrine manipulations or to chemotherapy than visceral or skin metastases, but this could reflect the relative insensitivity of our current methods of assessment more than true biological differences. Amongst the known inhibitors of osteoclast activity, bisphosphonates are the most promising drugs. Long-term administration of clodronate or Pamidronate to patients with breast cancer inhibits bone resorption and could reduce the occurrence of fractures, hypercalcemie episodes, and the development of new bone lesions. Objective sclerosis of osteolytic lesions has been obtained in patients receiving only bis-phosphonate therapy. If the initial results are confirmed, this new therapeutic modality will be one of the major recent therapeutic advances in clinical oncology.