The influence of dietary Ca on peak bone mass and on subsequent bone loss is controversial. Despite this an assessment of nutritional status is often included in the clinical evaluation of osteoporosis risk. To assess the value of this we investigated the relationship between current diet and bone mineral density (BMD) in 426 postmenopausal women, aged 45-59 years, who were enrolled into an international multi-centre trial of alendronate for the prevention of postmenopausal osteoporosis. BMD of the lumbar spine and proximal femur was measured on two occasions approximately 2 weeks apart by dual-energy X-ray absorptiometry. Serum osteocalcin was measured by immunoradiometric assay and serum 25-hydroxycholecalciferol by radioimmunoassay. Dietary assessment was performed by analysis of a 3 d unweighted dietary record, using Salford University's Microdiet software. BMD at both the lumbar spine and femoral neck correlated significantly with BMI, age, and average serum osteocalcin concentration. We therefore corrected for these variables in subsequent analyses. Dietary Ca intake ranged from 223 to 2197 mg/d (median 852 mg/d). Neither dietary Ca intake nor any other nutritional variable correlated significantly with BMD. There was a weak, but significant correlation between Ca intake and serum osteocalcin. We conclude that current diet does not correlate with BMD in early postmenopausal women. However, present diet may affect the rate of change of BMD, and this is supported by the finding of a significant relationship between dietary Ca and serum osteocalcin, a marker of bone formation rate.