To compare the effect of antihypertensive drugs on endothelium-dependent vasodilation in the peripheral conduit arteries of patients with essential hypertension, in a prospective, randomized, parallel group study, endothelial function was assessed in 168 hypertensive patients before and after 6-month treatment with randomly assigned nifedipine GITS (30 to 60 mg, n=28), amlodipine (5 to 10 mg, n=28), atenolol (50 to 100 mg, n=29), nebivolol (5 to 10 mg, n=28), telmisartan (80 to 160 mg, n=29), and perindopril (2 to 4 mg, n=28). If necessary, hydrochlorothiazide (25 mg) was added to each compound. We evaluated brachial artery flow-mediated, endothelium-dependent dilation (high-resolution ultrasound) compared with endothelium-independent response to glyceryl trinitrate (25 microg/s). Brachial artery diameter was measured by automatic computerized analysis. Forty healthy subjects were evaluated as a control group. Oxidative stress production was evaluated by measuring plasma malondialdehyde and plasma lipoperoxides; plasma antioxidant capacity was assessed as ferric-reducing antioxidant power. Hypertensive patients showed a significantly (P<0.01) lower flow-mediated dilation (5.2+/-1.9%) as compared with healthy control subjects (7.1+/-2.6%). Response to glyceryl trinitrate was similar in control subjects and patients. At baseline, blood pressure, diameter, flow-mediated dilation, and response to glyceryl trinitrate were similar in the different treatment groups. All treatments similarly reduced blood pressure, but only perindopril increased flow mediated dilation (from 5.1+/-2 to 6.4+/-2.4%; P<0.01) without modifying the response to glyceryl trinitrate. Perindopril but also telmisartan nifedipine and amlodipine reduced oxidative stress and increased plasma antioxidant capacity. In patients with essential hypertension, ACE inhibitors appear to be the only compounds able to improve conduit artery endothelium-dependent vasodilation.