OBJECTIVE: To test the hypothesis that the rate of degradation of exogenously administered albumin is faster with bolus administration than with continuous infusion and thus that a bolus administration is less efficacious in restoring blood albumin concentration (BAC) in the hypoalbuminemic critically ill pediatric patient. DESIGN: A prospective, controlled study of two groups of patients. SETTING: Pediatric intensive care unit (PICU) of a children's hospital. PATIENTS: 37 critically ill hypoalbuminemic patients (BAC < or = 2.8 g/dl), in whom no overt protein-losing disease was identified, were divided into two treatment groups and included in a 60-h study. INTERVENTIONS: 18 patients were given an i.v. bolus of 1 g/kg of 25% albumin over 4 h. This treatment was repeated after 24 and 48 h. Nineteen other patients were given the same dose of 1 g/kg of 25% albumin as a continuous 24-h infusion throughout the 60-h study period. BAC along with sodium, potassium, and total and ionized calcium were measured in the serum of blood samples obtained at predetermined intervals. MEASUREMENTS AND MAIN RESULTS: A 4 h bolus of albumin resulted in an acute rise in BAC, which declined to baseline within 24 h. A continuous infusion resulted in a steady rise in BAC with 24-h levels significantly higher than baseline. The percent change in mean BAC from baseline, calculated at 12-h intervals during the 60-h study period, showed a steady increase in the continuous infusion group with a 34% increase after the first 24 h. In contrast, the 4-h bolus method resulted in major fluctuations in the BAC values with only a 14% increase (p < 0.05) after 24 h. Albumin's volume of distribution, half-life and elimination constant, calculated based on blood albumin values during the first 24 h after the bolus administration, were 0.12 +/- 0.03 l/kg, 4.6 +/- 1.8 h and 0.17 +/- 0.06 h-1, respectively. This half-life did not apply to the continuous infusion group as a steady state was not achieved after 30 h (6 half-lives), and BAC continued to rise throughout the 60-h study period. No significant changes in blood electrolytes were observed with either method. CONCLUSIONS: The half-life of exogenous albumin in the critically ill hypoalbuminemic pediatric patient is short if given as a bolus. Continuous infusion therapy appears to be more efficacious in increasing BAC over time, as the half-life with this method appears to be longer.