Objective We analyzed the relationship between serum leptin levels and bone mineral density (BMD) values as well as the relationship between serum leptin levels and whole body composition, whether or not they were associated. In addition, we also investigated whether lean mass or fat mass is a better predictor of BMD in postmenopausal Turkish women. Design and measurements One hundred consecutive postmenopausal women with a mean age of 55.1 +/- 6.3 years who visited our outpatient clinic for the evaluation of osteoporosis were recruited. Skin fold thickness at four sites and waist:hip ratio were measured. Body mass index (BMI) was calculated in kg/m(2). Serum concentrations of leptin, insulin, and estradiol were evaluated. Bone formation and resorption markers were also determined. The BMD values were measured by dual energy X-ray absorptiometry (DEXA) at the lumbar spine and femoral neck. Whole body composition (lean mass, fat mass, and percentage of fat), total bone mineral content (BMC) in g, and total BMD were also measured by DEXA. Results Serum leptin levels did not correlate with BMD values at all skeleton sites measured. Leptin correlated positively with fat mass, percentage of fat, and BMI (r=0.738 and P=0.00, r=0.536 and P=0.00, r-0.356 and P=0.00, respectively). Lean mass correlated with BMD at all sites measured (r=0.339 and P=0.00, r=0.312 and P=0.01, r=0.523 and P=0.00, r=0.636 and P = 0.00). Lean mass correlated with BMI (r = 0.636, P = 0.00) but not with serum leptin (r =-0.021, P = 0.881), and it was an independent determinant of BMD at all skeleton sites measured. Conclusion Our study showed that lean mass is a better predictor than fat mass of bone mineral density and that serum leptin levels are not associated with BMD.