The influence of pregnancy on bone mineral density (BMD) was evaluated by dual-energy X-ray absorptiometry (DXA) in 73 women (mean age 29 years, range 20-44 years) postpartum. Fifty-five age-matched women served as controls. The influence of lactation was evaluated in 65 of the delivered women who were followed with repeated measurements, a mean of 4.5 +/- 0.1 and 11.5 +/- 0.1 months after the delivery. The influence of multiple pregnancies was evaluated in 39 premenopausal women (mean age 38 years, range 31-54 years) with a minimum of four pregnancies (range 4-7). Fifty-eight age-matched healthy premenopausal women with a maximum of two pregnancies (range 0-2) served as controls. Data are presented as mean +/- SEM. BMD data are adjusted for differences in total fat mass and total lean mass. Lumbar spine BMD was 7.6 +/- 0.1% and total body BMD 3.9 +/- 0.1% lower in women postpartum compared with controls (both p<0.001). BMD did not decrease significantly in non-breastfeeding mothers. Mothers breastfeeding for 1-6 months decreased femoral neck BMD by 2.0 +/- 1.0% during the first 5 months postpartum (p<0.001). No further BMD loss was seen between 5 and 12 months postpartum. Femoral neck BMD 12 months after delivery was 1.3 +/- 0.8% lower than after delivery in mothers breastfeeding for 1-6 months (p = 0.05). Mothers breastfeeding for more than 6 months decreased Ward's triangle BMD by 8.5 +/- 1.0% and lumbar spine BMD by 4.1 +/- 0.8% during the first 5 months postpartum (both p<0.05). No further BMD loss was seen between 5 and 12 months postpartum. Femoral neck BMD 12 months after delivery was 4.0 +/- 1.1% lower and Ward's triangle BMD 5.3 +/- 1.9% lower than after delivery in mothers breastfeeding for more than 6 months (both p<0.05). BMD loss was higher during the first 5 months following delivery in the lactating women compared with the non-lactating women (p<0.05 comparing lumbar spine BMD loss in lactating mothers versus non-lactating mothers). However, in women with a minimum of four pregnancies the BMD was no lower than in age-matched women with fewer pregnancies. Total duration of lactation was not correlated with the present BMD. In summary, pregnancy seem to confer a low BMD with additional BMD loss during 5 months of lactation. Even if complete restoration in BMD was not reached within 5 months of weaning, women with four pregnancies or more had a BMD no lower than women with two pregnancies or fewer. We conclude that neither an extended lactation period nor multiple pregnancies could be used as a risk factor when predicting women at risk for future osteoporosis.