Our objectives were 1) to determine whether exogenously administered arginine vasopressin (AVP) can exert a vasoactive influence on the pulmonary circulation of conscious dogs during specific vasopressinergic-1 (V1) receptor block, and 2) to assess the extent to which the pulmonary vascular response to AVP during V1 receptor block is mediated by either sympathetic beta-adrenergic or cholinergic receptor activation or by cyclooxygenase pathway activation. Multipoint pulmonary vascular pressure-cardiac index (P/Q) plots were constructed during normoxia in conscious dogs by stepwise constriction of the thoracic inferior vena cava to reduce Q. In dogs pretreated with a specific V1 receptor antagonist [d(CH2)5 AVP, 10 micrograms/kg iv], AVP infusion (7.6 ng.kg-1 X min-1 iv) increased (P less than 0.01) Q from 139 +/- 6 to 175 +/- 8 ml.min-1 X kg-1, and decreased (P less than 0.01) the pulmonary vascular pressure gradient (pulmonary arterial pressure-pulmonary capillary wedge pressure: PAP-PCWP) over the entire range of Q studied (140 to 80 ml.min-1 X kg-1). This pulmonary vasodilator response to AVP during V1 block was also observed following sympathetic beta-adrenergic block alone, following combined sympathetic beta-adrenergic and cholinergic block, and following cyclooxygenase pathway inhibition. Thus exogenous administration of AVP during specific V1 receptor block results in active, nonflow-dependent pulmonary vasodilation. This pulmonary vasodilator response is not mediated by reflex activation of sympathetic beta-adrenergic or cholinergic receptors or by metabolites of the cyclooxygenase pathway over a broad range of Q.