BLOOD FLOWS THROUGH THE CANINE HEPATIC (HBF) AND SUPERIOR MESENTERIC ARTERIES (MBF) WERE MEASURED WITH ELECTROMAGNETIC FLOWMETERS, DURING INFUSIONS OF VASOPRESSIN, BY THREE ROUTES OF ADMINISTRATION: 1) intra-hepatic-arterially (IHA), 2) intra-portal-venously (IPV) and 3) intra-systemic-venously (IV). Mean control HBF was 148 +/- 17 (S.E.) ml min(-1); MBF was 243 +/- 27 ml min(-1); aortic pressure (AP) was 126 +/- 3 mm Hg; portal venous pressure (PVP) was 8.8 +/- 1.0 mm Hg. Infusions of vasopressin, at a rate of 5 x 10(-3) units kg(-1) min(-1), IHA, reduced HBF significantly (p < .001) to 121 +/- 21 ml min(-1), within one minute. Flow returned to control, despite continued drug infusion; and at the end of the fifth minute of infusion, the value (134 +/- 21 ml min(-1)) was not significantly (p > .05) different from control. During the same infusion, MBF fell to 129 +/- 28 ml min(-1) (p < .001), by the sixth minute of the infusion and remained at this level for the duration of the infusion. AP increased to 137 +/- 13 mm Hg, by the sixth minute of the infusion and was sustained at this level for the duration of the infusion. PVP decreased to 7.0 +/- 1.0 mm Hg, by the tenth minute of the infusion. The responses to IPV vasopressin were indistinguishable from those following IHA vasopressin, with the exception that HBF was reduced to only 147 +/- 22 ml min(-1) (from a preinfusion control of 160 +/- 23 ml min(-1)), at one minute. HBF returned to control, despite continuation of the infusion. IV vasopressin, at the same concentration, caused no change in HBF throughout the ten minute infusion. These observations indicate that the canine hepatic arterial circulation responds to vasopressin with vasoconstriction characterized by autoregulatory escape. By any of the three routes of administration, vasopressin causes a significant and sustained reduction in blood flow through the superior mesenteric artery. Autoregulatory escape, from vasopressin-induced mesenteric arterial constriction, is not observed. Based on these observations, significant changes in mesenteric arterial blood flow can be anticipated without associated significant changes in hepatic arterial blood flow, regardless of the route of administration of vasopressin.