BACKGROUND Milrinone has been shown to increase cardiac output in children after cardiac surgery, but pharmacokinetic analysis has not been used to identify effective dose regimens. The purpose of this study was to characterize the pharmacokinetics of milrinone in infants and children and to apply the results to the issue of dosing. METHODS Twenty children were studied after they underwent repair of congenital cardiac defects. Control hemodynamic measurement was made after the children were separated from cardiopulmonary bypass, and each patient was given a loading dose of 50 microg/kg progressively in 5 min. Hemodynamic measurements were recorded again at the end of the loading dose and when a blood sample was taken to determine milrinone plasma concentrations. Further blood samples were taken during the next 16 h for milrinone plasma concentration analysis. The pharmacokinetics of milrinone were analyzed using the population pharmacokinetic program NONMEM. RESULTS The loading dose of milrinone resulted in a mean decrease in mean blood pressure of 12% and a mean increase in cardiac index of 18% at a mean peak plasma concentration of 235 ng/ml. The pharmacokinetics of milrinone were best described by a three-compartment model. In the optimal model, all volumes and distribution clearances were proportional to weight, and weight-normalized elimination clearance was proportional to age; ie., Cl1 = 2.5 x weight x (1 + 0.058 x age) where Cl1 is expressed as ml/min, and the units of weight and age are kg and months, respectively. CONCLUSIONS A loading dose of 50 microg/kg effectively increases cardiac index in children after cardiac surgery. Simulations indicate that the peak plasma concentration can be maintained by following the loading dose of 50 microg/kg with an infusion of approximately 3 microg x kg(-1) x min(-1) for 30 min and then a maintenance infusion, which may require adjustment for age.