We performed a prospective matched cohort study to investigate the effects of long-term (>1 month) heparin therapy on lumbar spine bone density. Twenty-five women who received heparin during pregnancy, and 25 matched controls underwent dual photon absorptiometry of the lumbar spine in the post-partum period. Zero of 25 heparin-treated patients developed fractures. Heparin-treated patients had a 0.082 g/cm2 lower bone density compared to untreated controls, which is clinically and statistically significant (p = 0.0077). There were 6 matched pairs in which only the heparin-treated patient had a bone density below 1.0 g/ cm2, compared to only one pair in which only the control patient had a bone density below this level (p = 0.089). The correlation coefficients of the difference .in bone density in each matched pair, and the duration of heparin therapy, the mean daily dose, and the total dose of heparin were 0.042, -0.015, and 0.021, respectively; none of these values is statistically significant. We conclude: 1) long-term heparin therapy was associated with a significant reduction in bone density, although fractures are uncommon, 2) there was no significant correlation between lumbar bone density and the dose or duration of heparin.