OBJECTIVE: To determine the factors that predict outcome of noninvasive ventilation (NIV) in critically ill children. DESIGN: Prospective observational study. SETTING: Multidisciplinary pediatric intensive care unit of a university hospital in Malaysia. PATIENTS: Patients admitted to the pediatric intensive care unit from July 2004 to December 2006 for respiratory support due to acute respiratory failure and those extubated from invasive mechanical ventilation. INTERVENTIONS: NIV was used as an alternative means of respiratory support for all children. In patients who had prior invasive mechanical ventilation, NIV was used to facilitate extubation, or it was used after a failed extubation. The children were assigned to the nonresponders group (intubation was needed) or responders group (intubation was avoided totally or for at least 5 days). The physiologic variables were monitored before, at 6 hrs, and 24 hrs of NIV. MEASUREMENTS AND MAIN RESULTS: Of 278 patients, 129 were admissions for management of acute respiratory failure and 149 patients received NIV to facilitate extubation (n = 98) or for a failed extubation (n = 48). Their median age and weight were 8.7 months (interquartile range, 3.1-33.1 months) and 5.5 kg (interquartile range, 3.3-10.8 kg), respectively. Intubation was avoided for > 5 days in 79.1% (n = 220). No significant difference in age or weight of responders and nonresponders was observed. The cardiorespiratory variables in all patients improved, but significant differences between the two groups were noted at 6 hrs and 24 hrs after NIV. CONCLUSIONS: NIV was a feasible strategy of respiratory support to avoid intubation in > 75% of children in this study. A higher Pediatric Risk of Mortality II score, sepsis at initiation of NIV, an abnormal respiratory rate, and a higher requirement of Fio2 may be predictive factors of NIV failure.