Background: Adequate oxygenation to treat hypoxia by suitable oxygen delivery pattern is the essence of pediatric critical care medicine. Monitoring oxygen saturation (SpO 2) is thus essential in order to decide course of treatment. It is commonly carried out using painless sequential non-invasive pulse oximetry as well by severely painful random arterial blood gas (ABG) analysis. Methods: This study compared both methods to determine the impact of various oxygen delivery patterns upon SpO 2 at bedside for meaningful and continuous evaluation. In a prospective cohort adopted for study, 60 children from acute pediatric wards of an academic tertiary referral hospital were observed. Children were eligible for enrollment using cross sectional approach, if the treating pediatricians ordered an ABG, on pulse oximetry monitoring with oxygen support by any oxygen delivery device using a validated checklist. Results: Among all the samples, 60 % were infants and 55 % were male. Complaints on admission were vomiting (35%), shortness of breath (33%), diarrhea (28%), lethargy (31%) and Kussmaul breathing (30%). SpO 2 was > 90% for 88% of children. Oxy hood box was used to deliver oxygen to 73% of children. SpO 2 had positive correlation with oxygen delivery methods (r = 0.8) and partial pressure of oxygen (PaO 2) by ABG analysis had positive correlation with SpO 2 by pulse oximetry (r = 0.9). Conclusions: Pulse oximetry can be used as a reliable bed side tool to evaluate oxygen delivery patterns and diagnose early respiratory failure in emergency settings where ABG analysis facility is not available.