This work was carried out in order to investigate possible relationships between bone turnover rate, as evaluated by bone biomarkers and skeletal mass, as evaluated by bone mineral density (BMD).Fifty-eight normal women and 30 female patients with osteoporotic fractures were enrolled. Three groups were defined: (1) fertile subjects (n = 24), mean age 33.7 +/- 8.1 years; (2) postmenopausal women (n = 32, including 11 patients with fractures) whose BMD values, in terms of T score, were less than -2.5 S.D. below the young adult mean obtained in our laboratory (mean age 61.7 +/- 7.9 years; and years since menopause (ysm), 12.6 +/- 8.3); (3) postmenopausal women (n = 32, including 19 patients with fractures) whose BMD values in terms of T score, were below -2.5 S.D. (mean age 62.9 +/- 8.6 years; and ysm 15.9 +/- 9.0). Groups II and III characterised, by inclusion criteria, by significant different mean BMD values, were similar as far as chronological and menopausal age were considered. Metabolic tests included a short urine collection to determine calcium, hydroxyproline, cross-linked N-telopeptides of type I collagen (NTx) and creatinine (Cr); half-way through this collection, a blood sample was taken for the measurement of total alkaline phosphatase activity (ALP) and tartrate-resistant acid phosphatase activity (TRAP). BMD at lumbar spine was evaluated.There were significant differences amongst the three groups in mean ALP (P0.001, by analysis of variance) TRAP (P0.006) and NTx/Cr (P0.001) values, but not as far as mean values of calcium/Cr or hydroxyproline/Cr ratios were concerned. Considering the group as a whole, there were significant inverse correlations between NTx/Cr, ALP, TRAP and BMD controlling for both age (r = -0.392, P0.001; r = -0.447, P0.001 and r = -0.327, P0.002, respectively) and ysm (r = -0.374, P0.001; r = -0.474, P0.001 and r = -0.333, P0.002).Our results indicate, that, even after controlling for both ageing and oestrogen status, there is an inverse relationship between bone mass (that at a given time represents the balance of all previous metabolic events) and a biochemical marker (which reflects bone turnover at the time of examination). These findings are in line with the belief that increased bone turnover should be regarded as a risk factor for osteoporosis. Furthermore, our results indicate that, unless there is no increase of hepatic isozyme, total ALP still maintains a possible role as a first analysis to evaluate bone turnover before requesting markers with greater specificity, sensitivity but also more expensive and whose analysis is sometimes time-consuming.