Women with the polycystic ovarian syndrome (PCOS) carry a number of cardiovascular risk factors, including insulin resistance, lipid abnormalities, and an altered pattern of sex steroid exposure. Noninvasive measurements of endothelial function, which can demonstrate abnormalities well in advance of clinically apparent disease, have not been previously reported in this patient group. We undertook a cross-sectional evaluation of endothelium-dependent and -independent vascular function using brachial artery ultrasound. We studied healthy women with clinical and laboratory evidence of PCOS (n = 18) and age-matched controls (n = 19), not taking any antihypertensive, cholesterol-lowering, or hormonal therapies. Laboratory parameters of insulin resistance, glycemia, cholesterol status, and hormone levels were also measured. Despite marked differences in glucose/insulin ratio [6.1 +/- 1.1 mmol/pmol (PCOS) vs. 9.9 +/- 0.6 (controls)] and free androgen index [11.9 +/- 2.3 (PCOS) vs. 3.7 +/- 0.6 (controls); normal, <5], we did not find evidence of impaired endothelial function in our patients with PCOS. Both endothelium-dependent (8.7 +/- 3.1%) and endothelium-independent (23.2 +/- 3.4%) vascular responses were normal, and practically identical to the responses seen in the control group (endothelium-dependent, 9.0 +/- 0.7; endothelium-independent, 23.0 +/- 1.2%). The PCOS women were more obese, but baseline brachial arterial diameters were not different between groups. There was no correlation between degree of insulin resistance or hyperandrogenism and the brachial response. This group of healthy obese young women with insulin resistance and hyperandrogenism due to PCOS had normal endothelium-dependent and -independent vascular responses compared to age-matched controls. The factors resulting in preservation of these response are unclear and warrant further investigation.