Objective: To delineate dobutamine pharmacokinetics and hemodynamic responses in children. Design: Prospective, pharmacokinetic study using sequential, graded dosing of drug. Interventions: Graded intravenous dobutamine infusions of 0.5, 2.5, 5, 10, and 20 mug/kg/min were sequentially administered for 25 mins each. Plasma dobutamine concentrations and echocardiographically determined hemodynamic data were obtained at baseline and at 15 and 25 mins during each infusion rate. Hemodynamic responses were evaluated by paired t-test and by computerized evaluation of individual dose-response curves. Setting. Pediatric intensive care unit in a university setting. Patients. Eleven stable, critically ill children previously requiring inotropic support with dobutamine. Seven patients were postcardiac surgical patients; four patients had acute cardiac dysfunction with septic shock and/or adult respiratory distress syndrome. Measurements and Main Results. Mean cardiac index increased from 3.8 to 5.2 L/min/m2 (p < .05). Increasing the infusion rate from 10 to 20 mug/kg/min increased cardiac index by 16% (p <.05). Cardiac index increased by >10% in four of seven patients at a dobutamine infusion rate of 0.5 mug/kg/min (mean 21%). The relationship of plasma dobutamine concentration to cardiac index, systolic blood pressure, and heart rate fit a threshold model with a log-linear relationship after the threshold in seven of nine, seven of 11, and eight of 11 patients, respectively. As anticipated, in the patients who responded, there were linear increases in hemodynamic responses with exponential increases in plasma dobutamine concentrations. Mean plasma clearance rate was 82 +/- 3 mL/min/kg. First-order kinetics were demonstrated by the direct linear relationship of plasma dobutamine concentration to infusion rate (mean r2 = .95; p < .01 for each patient) and by independence of clearance from dose and duration of each infusion. Conclusions: Dobutamine effectively improves systolic function in critically ill children. Hemodynamic responses to dobutamine generally follow a predicted log-linear dose-response model. Dobutamine clearance in this study was consistent with first-order kinetics. The wide variability in hemodynamic responses and clearance kinetics indicate that dobutamine infusions must be titrated individually.