Self-administration of anabolic-androgenic steroids to increase muscular strength and lean body mass has been used widely among athletes. Flow mediated dilatation (FMD) determined by ultrasound of the brachial artery is accepted as both an in vivo index of endothelial function and an indicator for future atherosclerosis. FMD was calculated in 20 male non-smoking body builders in different phases of their training cycle and in six male non-smoking control athletes. Ultrasound studies of the brachial artery were performed according to the protocol of Celermajer et al. Of the entire training cycle, work-out phase was training phase without actual intake of anabolic-androgenic steroids over 8 weeks; build-up phase included actual intake of anabolic-androgenic steroids; and competition phase consisted of 8 weeks post intake of anabolic-androgenic steroids. Baseline characteristics did not differ between body builder groups except for a higher weight in competition phase body builders. Hormonal analysis revealed suppressed luteinizing hormone and follicle stimulating hormone levels in build-up phase body builders. The lipid profiles showed a marked reduction of HDL-C in build-up phase body builders. FMD was reduced in body builders of all phases when compared to control athletes (work-out phase: 2.5+/-2.7%; build-up phase: 2.1+/-3.0%; competition phase: 0.4+/-2.9% vs. 10.9+/-4.4%, P<0.05 by pairwise comparison using Scheffe's test for work-out phase, build-up phase and competition phase vs. control athletes). The glyceryl trinitrate-induced vasodilatation was diminished, though not statistically significantly, in body builders when compared with control athletes. The differences in FMD persisted after adjustment for vessel size. Our data indicate that intake of anabolic-androgenic steroids is associated with both an atherogenic blood lipid profile and endothelial dysfunction and thus may pose an increased risk of atherosclerosis.