Pre-clinical data have shown that osteoprotegerin (OPG) inhibits osteoclast function and therefore plays an important role in bone remodelling. This study aimed to evaluate the clinical value of serum OPG. Do higher OPG serum levels reflect decreased bone resorption and perhaps higher bone mass in women? Serum OPG levels were measured in 177 healthy women (aged 17-85 years) and in 48 untreated patients (mean age 71 +/- 5) with established osteoporosis, and related to age, bone mass, markers of bone turnover and, in the case of patients with osteoporosis, to pre-existing vertebral fractures. In healthy women OPG levels showed a positive correlation with age (r = 0.25, p < 0.001) but not to bone mass or markers of bone turnover. In women with osteoporosis, however, there was a strong relationship between serum OPG and markers of bone turnover (serum c-terminal crosslinked telopeptides of thpe I collagen (sCTX): r = +0.82, p < 0.0001; osteocalcin (OC): r = +0.69, p < 0.0001), with patients who had higher levels of bone-turnover markers showing higher serum levels of OPG. After adjustment for bone mass and bone markers, patients with pre-existing vertebral fractures had significantly lower serum OPG levels than patients without fractures (57 +/- 8 vs. 97 +/- 10 pg/ml, [mean +/- SE], p < 0.01). The age-dependent increase of OPG as an antiresorptive factor may reflect an insufficient paracrine mechanism of bone cells to compensate for bone loss in older age. In patients with osteoporosis, however, OPG correlated strongly with markers of bone turnover; this may point toward a higher level of RANKL/OPG expression in these patients. Finally, low OPG serum levels seem to be associated with vertebral fractures. We hypothesise that low OPG levels in preset conditions of bone turnover may indicate a higher risk of fracture in patients with osteoporosis.