1. It was the aim of the present study to investigate the role of the beta-adrenergic receptor in hypoxaemia induced vasodilatation in the human forearm. 2. The study was performed in 12 non-smoking male volunteers. In six subjects the local vascular effects of intra-arterially (i.a.) infused propranolol (0.1 mu kg-1 min-1) was determined during normoxaemia and hypoxaemia (peripheral oxygen saturation; SpO2 80%), and compared with the contra-lateral (control) arm. beta-adrenergic receptor blockade by propranolol was confirmed by i.a. infusions of adrenaline. In six other subjects the effects of incremental hypoxaemia (SpO2 90, 85, 80%) on forearm- and finger blood flow was investigated. A difference between these vascular beds is the absence of vascular beta-adrenergic receptors in the finger. Forearm- and finger blood flow were measured by venous occlusion plethysmography. Plasma levels of (nor-)adrenaline were determined in both arterial and venous blood samples. Hypoxaemia was attained by gradual decompression of a hypobaric chamber and individually adjusted. 3. During normoxaemia the single infusion of propranolol did not influence forearm vascular resistance. In contrast, during hypoxaemia a net vasoconstriction was observed in the arm treated with propranolol, which was significantly different (P = 0.009) from the vasodilatation in the control arm (mean difference in response 40%; 95% confidence interval 13.2-66.8). The net arterio-venous spillover of noradrenaline from the forearm increased after the first 15 min of hypoxaemia (P < 0.05) and returned to baseline in the next 15 min. Arterial and venous plasma levels of adrenaline during hypoxaemia remained unchanged compared with normoxaemia. In the second set of experiments incremental levels of hypoxaemia induced a vasoconstriction in the finger, which was significantly different (P = 0.025) from the vasodilatation in the forearm (mean difference in response 300%; 95% confidence interval 63-537). 4. The data indicate that beta-adrenergic receptors contribute to hypoxaemia induced vasodilatation, despite unaltered adrenaline plasma concentrations.