This is a retrospective review of children 15 years of age or younger, who underwent overnight sleep studies between 1980 and 1993. All were diagnosed and treated for obstructive sleep apnea (OSA). Overnight studies were performed for OSA in 413 children. One hundred seventy-five (42.4%) children were treated with adenotonsillectomy and 80 (19.4%) with nasal mask continuous positive airway pressure (nCPAP). The proportion of male children was greater than expected in both the entire study group (69%, p < 0.001) and in those treated with nCPAP for OSA (71% p < 0.001). There was no significant difference between the mean age of the children treated with nCPAP (5.7 +/- 0.5 yr) and the entire group studied (5.04 +/- 0.21 yr). A greater proportion of the children who received nCPAP therapy had a congenital syndrome or malformation than in the group with OSA as a whole; 27.7% of children assessed for OSA were affected, and 53% of those children with OSA who received treatment with nCPAP (p < 0.001). Therapy with nCPAP (mean duration 15 +/- 3 mo, mean pressure 7.9 cm H2O) eliminated the signs of OSA in 72 children (90%). Respiratory disturbance index fell from a mean of 27.3 +/- 20.2 to 2.55 +/- 2.74 (p < 0.001). Eight of 32 children who underwent pressure determination studies could not tolerate nCPAP above an upper limit because of hypoventilation or frequent central apneas. Nevertheless, we conclude that nCPAP is an effective and generally well-tolerated therapy for treatment of OSA in infants and children.