Fluid resuscitation of burned children is challenging because of their small size and intolerance to over- or underresuscitation. Our American Burn Association-verified regional burn center has used colloid “rescue” as part of our pediatric resuscitation protocol. With Institutional Review Board approval, the authors reviewed children with ≥15% TBSA burns admitted from January 1, 2004, to May 1, 2009. Resuscitation was based on the Parkland formula, which was adjusted to maintain urine output. Patients requiring progressive increases in crystalloid were placed on a colloid protocol. Results were expressed as an hourly resuscitation ratio (I/O ratio) of fluid infusion (ml/kg/%TBSA/hr) to urine output (ml/kg/hr). We reviewed 53 patients; 29 completed resuscitation using crystalloid alone (lactated Ringer's solution [LR]), and 24 received colloid supplementation albumin (ALB). Groups were comparable in age, gender, weight, and time from injury to admission. ALB patients had more inhalation injuries and larger total and full-thickness burns. LR patients maintained a median I/O of 0.17 (range, 0.08–0.31), whereas ALB patients demonstrated escalating ratios until the institution of albumin produced a precipitous return of I/O comparable with that of the LR group. Hospital stay was lower for LR patients than ALB patients (0.59 vs 1.06 days/%TBSA, P = .033). Twelve patients required extremity or torso escharotomy, but this did not differ between groups. There were no decompressive laparotomies. The median resuscitation volume for ALB group was greater than LR group (9.7 vs 6.2 ml/kg/%TBSA, P = .004). Measuring hourly I/O is a helpful means of evaluating fluid demands during burn shock resuscitation. The addition of colloid restores normal I/O in pediatric patients.