In cats anesthetized with pentobarbital, isoproterenol infused into a peripheral vein causes a reduction in hepatic blood volume measured by plethysmography. As this response is accompanied by increases in portal and hepatic lobar venous pressures, the decrease in hepatic volume cannot be a passive emptying secondary to reduced intrahepatic pressure. We conclude that intravenous isoproterenol causes an active hepatic venoconstriction. Nifedipine produced similar responses. From this and our previous data, we conclude that in anesthetized cats, arteriolar vasodilators which increase cardiac output cause hepatic venoconstriction (hydralazine, adrenaline, dopamine, isoproterenol, and nifedipine), while those which do not increase cardiac output have no effect on the hepatic venous bed (nitroprusside and diazoxide) or cause venodilatation (nitroglycerine). The mechanism of the hepatic venoconstrictor effect of isoproterenol was investigated further. Because previous work has shown that this response does not occur when isoproterenol is infused locally into the hepatic artery or portal vein, the venoconstrictor effect of peripheral intravenous infusions must be indirectly mediated. The response was still present after hepatic denervation, adrenalectomy, nephrectomy, and after indomethacin administration indicating it is not mediated by the hepatic nerves, adrenal catecholamines, the renal renin-angiotensin system, or prostaglandins. The mechanism remains unknown.