High plasma concentrations of N-acetylamrinone, a primary metabolite of amrinone, are measured in some children during prolonged amrinone infusion. The purpose of this investigation was to determine if N-acetylamrinone has direct hemodynamic effects independent of amrinone. Twenty neonatal piglets received an infusion of 6 x 10(9) colony-forming units/kg of group B Streptococcus to induce sepsis. Subsequently, they were divided into 1 of 3 groups and received a 1-hr infusion of either normal saline (N = 4); 8 mg/kg amrinone, followed by 20 micrograms/kg/min (N = 9); or 8 mg/kg N-acetylamrinone, followed by 20 micrograms/kg/min (N = 7). Hemodynamic measurements and arterial/venous blood-gas determinations were obtained every 30 min during the study. Systemic vascular resistance and pulmonary vascular resistance were calculated. One milliliter of blood was obtained every 30 min during drug administration to determine plasma amrinone and N-acetylamrinone concentrations. The mean amrinone plasma concentrations measured at 30 and 60 min during the infusion time in the group receiving amrinone were 8.8 +/- 1.1 and 6.9 +/- 0.7 micrograms/ml, respectively. These animals experienced a significant decrease in mean pulmonary artery pressure and pulmonary vascular resistance, compared with saline controls after a 30-min infusion of amrinone. The mean N-acetylamrinone plasma concentrations measured at 30 and 60 min during the N-acetylamrinone infusion were 7.3 +/- 0.8 and 5.7 +/- 0.6 micrograms/ml, respectively. There was no difference between any hemodynamic parameter measured in these animals, compared with saline controls at any time during the infusion. We conclude that amrinone, but not N-acetylamrinone, causes pulmonary vasodilation in a porcine model of sepsis and that the parent drug is the sole active component in amrinone.