Although red wine (RW) reduces cardiovascular risk, the mechanisms underlying the effect have not been identified. Correction of endothelial dysfunction by RW flavonoids could be one mechanism. We measured brachial artery reactivity by high-resolution ultrasonography, plasma lipids, glucose, adhesion molecules (ICAM-1 and VCAM), and platelet function in 16 hypercholesterolemic individuals (8 men and 8 women; mean age 51.6 +/- 8.1 years) without other risk factors. Twenty-four normal subjects were used as controls for vascular reactivity. Subjects randomly received RW, 250 ml/day, or purple grape juice (GJ), 500 ml/day, for 14 days with an equal wash-out period. At baseline, all 16 subjects were hypercholesterolemic (mean LDL = 181.0 +/- 28.7 mg/dl) but HDL, triglycerides, glucose, adhesion molecules, and platelet function were within normal limits. Brachial artery flow-mediated dilation was significantly decreased compared to controls (9.0 +/- 7.1 vs 12.1 +/- 4.5%; P < 0.05) and increased with both GJ (10.1 +/- 7.1 before vs 16.9 +/- 6.7% after: P < 0.05) and RW (10.1 +/- 6.4 before vs 15.6 +/- 4.6% after; P < 0.05). RW, but not GJ, also significantly increased endothelium-independent vasodilation (17.0 +/- 8.6 before vs 23.0 +/- 12.0% after; P < 0.01). GJ reduced ICAM-1 but not VCAM and RW had no effect on either molecule. No significant alterations were observed in plasma lipids, glucose or platelet aggregability with RW or GJ. Both RW and GJ similarly improved flow-mediated dilation, but RW also enhanced endothelium-independent vasodilation in hypercholesterolemic patients despite the increased plasma cholesterol. Thus, we conclude that GJ may protect against coronary artery disease without the additional negative effects of alcohol despite the gender.