© 2019 Elsevier Inc. Objective: To determine the impact of policy changes for pulse oximetry oxygen saturation (SpO2) alarm limits on neonatal mortality and morbidity among infants born very preterm. Study design: This was a retrospective cohort study of infants born very preterm in the Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network. Infants were classified based on treatment at a hospital with an SpO2 alarm policy change and study epoch (before vs after policy change). We used a generalized linear mixed model to determine the effect of hospital group and epoch on the primary outcomes of mortality and severe retinopathy of prematurity (ROP) and secondary outcomes of necrotizing enterocolitis, bronchopulmonary dysplasia, and any ROP. Results: There were 3809 infants in 10 hospitals with an SpO2 alarm policy change and 3685 infants in 9 hospitals without a policy change. The nature of most policy changes was to narrow the SpO2 alarm settings. Mortality was lower in hospitals without a policy change (aOR 0.63; 95% CI 0.50-0.80) but did not differ between epochs in policy change hospitals. The odds of bronchopulmonary dysplasia were greater for hospitals with a policy change (aOR 1.65; 95% CI 1.36-2.00) but did not differ for hospitals without a policy change. Severe ROP and necrotizing enterocolitis did not differ between epochs for either group. The adjusted odds of any ROP were lower in recent years in both hospital groups. Conclusions: Changing SpO2 alarm policies was not associated with reduced mortality or increased severe ROP among infants born very preterm.