We first discuss the question of direct versus indirect mechanisms in the vasoconstrictor effects induced by low plasma levels of arginine vasopressin (AVP). By infusing AVP directly into the arteries supplying various organs and tissues and measuring regional blood flows with radioactive microspheres in conscious dogs, it was determined that the increase in resistance measured after systemic administration of small amounts of AVP in the skeletal muscle, small intestine, colon, and abdominal fat results in large part from indirect mechanisms. Direct vasoconstrictor effects of AVP at these plasma concentrations were limited to the skin, the pancreas, and the compact bones. We then consider the question of cardiovascular effects of AVP from its antidiuretic activity. In the presence of a vascular antagonist, AVP decreased peripheral resistance and increased cardiac output in conscious dogs. An analog with selective antidiuretic activity, VDAVP, produced the same effects as the combination of vasopressin plus vascular antagonist. These data indicate that vasopressin, by its antidiuretic activity, produces cardiovascular effects that are opposed to many of those produced by its vasoconstrictor action. Studies in 48-h dehydrated dogs showed that part of the hemodynamic response to blockade of the vasoconstrictor action of vasopressin under these conditions is caused by unmasking cardiovascular effects linked to the antidiuretic activity of the arginine vasopressin molecule.