To generate quantitative data relating to the potential hypertensive activity of arginine vasopressin (AVP), 140 and 560 microunits AVP/kg/min were infused chronically in both normotensive dogs and dogs made hypertensive by chronic infusion of either angiotensin II (AII) or aldosterone. The lower rate of AVP infusion increased plasma AVP concentration from 0.4 +/- 0.1 microunits/ml. Mean arterial pressure (MAP) was recorded 24 hours per day, and all dogs were infused continuously with 800 ml of isotonic saline per day. During the initial days of AVP infusion in normotensive dogs, natriuresis, kaliuresis, and water retention were prominent and MAP increased progressively to a peak on Day 6 (30 mm Hg above control). Subsequently, diuresis ensued, net water retention decreased, and MAP fell progressively to only 13 mm Hg above control by Day 12 of AVP infusion. In contrast, in AII or aldosterone hypertensive dogs during AVP infusion, the natriuresis was greatly attenuated, water balance was unchanged or even negative, and MAP either did not increase or increased only transiently. When AVP infusion was terminated in dogs given only AVP, diuresis occurred, and MAP fell gradually over a period of hours to hypertensive levels. In marked contrast, cessation of AVP infusion in dogs with either AII or aldosterone hypertension was associated with a precipitous fall in MAP of 35 to 40 mm Hg within 1 hour: however, this reduction was only transient - over the subsequent hours, both salt and water retention occurred, and MAP returned to previous hypertensive levels. Thus, in the hypertensive models studied, high plasma levels of AVP had relatively weak hypertensive effects. Although variations in plasma AVP concentration were associated with rather pronounced acute effects on MAP, the long-term changes in MAP produced by AVP were either minimal or, in the case of the animals made hypertensive by other agents, nonexistent.