1 The alleged importance of cardiac f,-adrenoceptors for the baroreceptor induced rise in cardiac output after acute vasodilatation was assessed in 41 patients with essential hypertension. 2 Diazoxide (300 mg i.v.) was given, to patients 1) when untreated (n = 29), 2) during treatment with propranolol (320 mg/day), (n= 15), or 3) during propranolol plus atropine (0.04 mg/kg), (n= 12).3 Diazoxide-induced reductions in arterial pressure during propranolol, either alone (-23 + 3%) or combined with atropine (-22 + 3%), were not significantly different from those without pretreatment (-24 + 3%, mean + s.e. mean. 4 The response of heart rate to diazoxide was somewhat diminished during propranolol (+ 14 + 2 with propranolol v +21 + 3% without propranolol, 15 paired observations, P < 0.001).5 Stroke volume rose more in response to diazoxide after pretreatment with propranolol (+ 16 + 11 with propranolol v + 2 + 5% without propranolol, P < 0.001) so that the response of cardiac output was not altered by ,B-adrenoceptor blockade (+ 32+4 with propranolol v + 24 + 9% without propranolol, P > 0.05).6 The rise in cardiac output was markedly diminished by additional parasympathetic blockade (+ 14+ 5% with propranolol plus atropine, n = 12, v 32 +4% with propranolol alone, n = 15, P < 0.01). 7 Increments in plasma noradrenaline were not significantly different in the three situations, indicating that baroreceptor sensitivity was not altered. 8 We conclude that the baroreflex induced rise in cardiac output during vasodilator treatment of hypertension depends on withdrawal of parasympathetic tone rather than sympathetic stimulation.