Besides an evident diuretic effect, amiloride has been shown to exert direct vasoactivity in various animal experiments, whereas human data on this issue are lacking. Inhibition of Na+/H+ exchange, alpha-adrenergic blockade, and sodium and calcium channel antagonism have been proposed as possible mechanisms of this action. Although the role of Na+/H+ exchange in vascular-tone modulation is not completely clear, various vasoconstrictive agents (e.g., angiotensin II) enhance its activity. We examined the direct effects of amiloride on human arterial vasculature in vivo. Forearm vasodilator responses to the infusion of placebo and amiloride (n = 10; 0.1-100 microg/min/dl) into the brachial artery were recorded by venous occlusion strain-gauge plethysmography. Reduction of forearm blood flow after local administration of noradrenaline or angiotensin II was measured before and after local amiloride administration. Amiloride increased the ratio of the infused/ noninfused forearm blood flow at the highest dosages (10, 30, and 100 microg/min/dl with 14+/-9, 17+/-14, 58+/-23% (p = 0.002, repeated-measures analysis of variance). In contrast to noradrenaline-induced vasoconstriction, the vasoconstrictor response to angiotensin II was significantly attenuated by amiloride (p = 0.02). At high concentrations, amiloride exerts direct vasodilator activity in human arterial vasculature in vivo. This effect appears not to depend on alpha-adrenergic receptor blockade, but shows interaction with angiotensin II, an activator of Na+/ H+ exchange.