In humans, hypoxia leads to increased sympathetic neural outflow to skeletal muscle. However, blood flow increases in the forearm. The mechanism of hypoxia-induced vasodilation is unknown. To test whether hypoxia-induced vasodilation is cholinergically mediated or is due to local release of adenosine, normal subjects were studied before and during acute hypoxia (inspired O(2) 10.5%; approximately 20 min). In experiment I, aminophylline (50-200 microg. min(-1). 100 ml forearm tissue(-1)) was infused into the brachial artery to block adenosine receptors (n = 9). In experiment II, cholinergic vasodilation was blocked by atropine (0.4 mg over 4 min) infused into the brachial artery (n = 8). The responses of forearm blood flow (plethysmography) and forearm vascular resistance to hypoxia in the infused and opposite (control) forearms were compared. During hypoxia (arterial O(2) saturation 77 +/- 2%), minute ventilation and heart rate increased while arterial pressure remained unchanged; forearm blood flow rose by 35 +/- 6% in the control forearm but only by 5 +/- 8% in the aminophylline-treated forearm (P < 0.02). Accordingly, forearm vascular resistance decreased by 29 +/- 5% in the control forearm but only by 9 +/- 6% in the aminophylline-treated forearm (P < 0.02). Atropine did not attenuate forearm vasodilation during hypoxia. These data suggest that adenosine contributes to hypoxia-induced vasodilation, whereas cholinergic vasodilation does not play a role.