Objective To describe and analyze the characteristics and the early risk factors for mortality of noninvasive ventilation (NIV) in critically ill children. Study design A multicenter, prospective, observational 2‐year study carried out with critically ill patients (1 month – 18 years of age) who needed NIV. Clinical data and NIV parameters during the first 12 h of admission were collected. A multilevel mixed‐effects logistic regression was performed to identify mortality risk factors. Results A total of 781 patients (44.2 ± 57.7 months) were studied (57.8% male). Of them, 53.7% had an underlying condition, and 47.1% needed NIV for lower airway respiratory pathologies. Bi‐level NIV was the initial support in 78.2% of the patients. Continuous positive airway pressure (CPAP) was used more in younger patients (33.7%) than in older ones (9.7%; p < .001). About 16.7% had to be intubated and 6.2% died. The risk factors for mortality were immunodeficiency (odds ratio OR = 11.79; 95% confidence interval CI = 2.95–47.13); cerebral palsy (OR = 5.86; 95% CI = 1.94–17.65); presence of apneas on admission (OR = 5.57; 95% CI = 2.13–14.58); tachypnea 6 h after NIV onset (OR = 2.59; 95% CI = 1.30–6.94); and NIV failure (OR = 6.54; 95% CI = 2.79–15.34). Conclusion NIV is used with great variability in types of support. Younger children receive CPAP more frequently than older children. Immunodeficiency, cerebral palsy, apneas on admission, tachypnea 6 h after NIV onset, and NIV failure are the early factors associated with mortality.