The purpose of this study was to analyze the ability of sera to reflect the state of bone metabolism by testing the osteogenic response of mesenchymal cells in culture. Sera of 20 peri- and postmenopausal women were tested before the initiation of hormone replacement therapy. The responding cells were osteoprogenitors (OPC) of rat marrow stroma which normally respond to dexamethasone (DEX) and β-glycerophosphate (βGP) by proliferation, differentiation, and mineralization in culture. Instead of DEX, diluted sera (1:50) were applied to rat stromal cell cultures for analysis of their ability to affect cell proliferation, specific alkaline phosphatase (ALP) activity, and cell-mediated mineralization. The results were compared individually with the respective values of vertebral bone mineral density (BMD), expressed as the number of standard deviations above or below the mean BMD of reference populations (positive or negative Z-score). Serum donors were divided in two; the group with positive Z-scores was considered to have a low risk, and that with negative Z-scores was considered to have a higher risk for vertebral fractures. No significant difference was found between the two groups in the ability of their sera to induce cell proliferation or specific ALP activity. However, sera representing negative Z-scores induced sixteenfold less mineralization than those of positive Z-scores. The scatter of individual mineralization values was highly discriminatory between the two groups (α < 0.00). These results indicate that the serum-induced, cell-mediated mineralization in culture might be suitable for initial evaluation of fracture risk and thus deserve further investigation. © 1996 Wiley-Liss, Inc.