Few studies have assessed autotitrating positive airway pressure (autoPAP) for treatment of obstructive sleep apnea (OSA) in children. We aimed to review our use of autoPAP for initiation of continuous positive airway pressure (CPAP) therapy in children, and compare autoPAP-derived treatment pressures to CPAP treatment pressure determined by attended polysomnography (PSG).Retrospective review of children initiated on autoPAP from 2013 to 2015. Mean autoPAP pressure (AutoMean pressure) and average device pressure ≤ 90% of time (Auto90 pressure) were taken from downloaded data and compared to the recommended treatment pressure following titration PSG (PSG pressure).Fifty-two children started CPAP, of whom 26 (age ± standard deviation 11.9 ± 3.4 years) used autoPAP and had titration PSG. AutoPAP was used on average 84% of nights (standard deviation 20%) in the first month, with a mean ± standard deviation 6.3 ± 2.0 hours of use on nights used. The median (interquartile range) obstructive apnea-hypopnea index decreased from 16.6 (11, 35) events/h before treatment to 2.2 (0.4, 3.8) events/h on the titration PSG. Median (interquartile range) PSG pressure was 9.0 cm H2O (7.0, 10.0), AutoMean pressure was 6.3 cm H2O (5.3, 7.5), and Auto90 pressure was 8.1 cm H2O (7.1, 9.5). These were significantly different (P < .001), with the significant difference lying between AutoMean and the other two pressures. PSG pressure was greater than or equal to the AutoMean pressure in all cases, and greater than or equal to the Auto90 pressure in 20 out of 26 cases (77%).AutoPAP is a safe and effective means of initiating CPAP in children. AutoMean and Auto90 pressures are usually below treatment pressure determined by titration PSG.