The development of bone metastases in cancer can be monitored easily using three markers: 24 h urinary hydroxyproline excretion (HOP) (an index of osteoclastic activity), serum alkaline phosphatase (Alk.Ph.) (an index of osteoblastic activity) and 24 h whole body retention of 99mTc-methylene diphosphonate (WBR%) (an index of bone turnover). To evaluate the effectiveness of this group of bone tumor markers in breast cancer we compared it with the following group of three markers which are commonly used in the monitoring of breast cancer and in the follow-up of advanced disease with or without bone metastases: carcinoembryonic antigen (CEA), tissue polypeptide antigen (TPA) and breast carcinoma antigen (CA 15/3). In 48 patients with bone metastases CEA, TPA and CA 15/3 were shown to be sensitive (79%, 85%, 90% respectively), while HOP, Alk.Ph. and WBR%, which are commonly accepted as reliable markers of bone activity, showed a lower sensitivity (67%, 46%, 75% respectively). These results may be explained by the lack of osteoclastic or osteoblastic (or both) activity at the time of diagnosis. This explanation is supported by the fact that the bone markers HOP, Alk.Ph. and WBR% were found to be more sensitive than the others in the subsequent follow-up study. We conclude that in our study, CEA, TPA and CA 15/3 are at first more sensitive than Alk.Ph., HOP and WBR% but during the follow-up Alk.Ph., HOP and WBR% are possibly both more specific and more sensitive.