Hormone replacement therapy (HRT) may reduce atherosclerosis among postmenopausal women, partly by reducing vascular endothelium damage. We have tested this hypothesis by evaluating the association of HRT with firstly, carotid intima media thickness (IMT) and plaques, and secondly, with endothelial cell damage, indicated by soluble thrombomodulin (sTM). Then, we tested the association between the two markers of atherosclerosis and the levels of sTM. Among 747 postmenopausal women included into the EVA study, we compared 154 HRT users (including 80% transdermal treatment) with 593 never users. Carotid IMT and plaques were measured with B-mode ultrasonography and sTM with ELISA. At least one plaque was detected among 13.6% of HRT users and 27.3% of never users. After adjustment for confounding factors, the odds ratio for the presence of plaque was 0.45 (95% confidence interval, 0.25-0.78, P=0.005) in HRT users in comparison with nonusers. HRT users had a slightly lower crude mean IMT than nonusers, but the difference was not significant. sTM was positively associated with mean IMT (P for trend=0.001) but not with plaques. Finally, estrogen users had a lower sTM level than nonusers (difference 0.14 ng/ml, P=0.03). As HRT was associated with sTM and plaques, but not with IMT, while sTM was only associated with IMT, our hypothesis was not confirmed. This suggests that the possible beneficial effects of HRT on atherosclerosis may not go through the endothelial cell damage assessed by plasma thrombomodulin.