A prospective, randomized trial was performed to determine whether maintaining serum albumin levels in burned pediatric patients had any effect on morbidity and mortality. Patients < 19 years of age with burns > 20% total body surface area were randomized to receive supplemental albumin to maintain levels 2.5 to 3.5 g/dL ("High Albumin") or were given albumin only if levels dropped < 1.5 g/dL ("Low Albumin") after completing burn shock resuscitation. The 36 patients in the Low Albumin group were well matched for age, burn size, depth of injury, and inhalation injury when compared with the High Albumin group (34 patients). As expected, serum albumin levels were significantly lower in the Low Albumin group when compared with the High Albumin group. No differences between groups were noted for resuscitation needs, maintenance fluid requirements, urine output, tube feedings received, days of antibiotic treatment, or ventilatory requirements. No differences in hematology, electrolytes, or nutritional laboratories were found. Finally, length of stay, complication rate, and mortality were not affected by albumin treatment. Albumin supplementation to maintain normal serum levels does not seem to be warranted in previously healthy children who suffer severe burns and who receive adequate nutrition.