1 The hepatic portal vein of the anesthetized dog was cannulated and perfused with blood derived from the cannulated superior mesenteric vein. 2 The portal vein was perfused at constant flow, the hepatic portal venous pressure being monitored continuously together with the inferior vena caval pressure. From these measurements, the hepatic portal venous vascular resistance was calculated. 3 Noradrenaline and adrenaline were injected intraportally in graded doses which caused dose-dependent increases in the hepatic portal vascular resistance. At all doses, adrenaline was significantly (P less than 0.05) more potent than noradrenaline. 4 Intraportal injections of vasopressin caused reductions in calculated hepatic portal venous vascular resistance in most experiments; three effects were dose-dependent. 5 No tachyphylaxis to the effects of noradrenaline, adrenaline or vasopressin was observed. 6 Intraportal injections of angiotensin caused dose-dependent increases in calculated hepatic portal vascular resistance up to 5 mug; therafter larger doses caused smaller increases in portal resistance. 7. Repeated intraportal injections of angiotensin revealed the existence of tachyphylaxis in the hapatic portal vascular bed. 8 Intraportal infusions of anagiotensin caused rises in calculated hepatic portal vascular resistance from which there was almost complete 'escape' despite the continued infusions. Infusions of noradrenaline which caused similar rises in calculated portal vascular resistance did not exhibit equivalent degrees of 'escape'. 9 The development of tachyphylaxisx explains the fact that doses of 10 and 20 mug of angiotensin injected after 5 mug doses produced smaller effects. If a much longer time interval was allowed between injections (30 min), the dose-response curve to angiotensin had a sigmoid shape. 10 These findings are discussed with respect to their possible importance in the functional status of the hepatic portal vascular bed in this species.