OBJECTIVE To determine to what extent the risk of bronchopulmonary dysplasia is affected by ventilatory management before the first dose of rescue artificial surfactant. METHODS Retrospective cohort study. METHODS One hundred eighty-eight low-birth-weight infants (< or = 1700 g) who received artificial surfactant therapy for respiratory distress syndrome and who were alive at 36 weeks of gestational age. RESULTS Bronchopulmonary dysplasia was defined by a need for supplemental oxygen to maintain an arterial saturation of 92% or more at 36 weeks of gestational age. RESULTS Thirty-seven percent (70/188) of the cohort met study criteria for bronchopulmonary dysplasia. Early determinants significantly associated with bronchopulmonary dysplasia (given as odds ratio, 95% confidence interval) in the most parsimonious backward stepwise logistic regression model included the following: birth weight of 1000 g or less (5.1, 2.4 to 10.7), cesarean birth because of fetal distress (4.4, 1.7 to 11.4), ventilatory efficiency index of 0.15 or less before surfactant therapy (3.1, 1.4 to 6.8), arterial-alveolar oxygen ratio of 0.15 or less before surfactant therapy (2.2, 1.01 to 4.6), and a low arterial PCO2 (< or = 29 vs > or = 40 mm Hg, 5.6, 2.0 to 15.6; 30 to 39 vs > or = 40 mm Hg, 3.3, 1.3 to 8.3). The inverse relationship between hypocarbia and bronchopulmonary dysplasia persisted even in stratified analyses limited to infants with measures of cardiovascular or respiratory illness that suggested less severe manifestations of disease. CONCLUSIONS Ventilatory management before rescue treatment with artificial surfactant therapy that result in hypocarbia may increase the risk of bronchopulmonary dysplasia. These findings suggest that early ventilatory management should not only provide adequate oxygenation but also limit hyperventilation.