BackgroundEfficacy of antenatal corticosteroids before 25 weeks of gestation is unclear.ObjectiveTo assess and compare neonatal outcomes following ANC exposure at 22, 23 and 24 weeks of gestation by conducting systematic review and meta- analysis.MethodsA systematic review of randomised controlled trials (RCT) and non-RCTs reporting on neonatal outcomes after exposure to ANC up to 246 weeks of gestation using the Cochrane systematic review methodology. Databases Pubmed, CINAHL, Embase, Cochrane Central library, and online abstracts of conference proceedings including the Pediatric Academic Society (PAS) were searched in Feb 2017. Primary outcome was in-hospital mortality defined as death before discharge during the first admission. Secondary outcomes included severe intraventricular hemorrhage (IVH> grade III and IV)/or periventricular leukomalacia (PVL), necrotising enterocolitis (NEC >stage II) and chronic lung disease (CLD). Meta-analysis was performed using a random-effects model. The level of evidence (LOE) was summarised using the GRADE guidelines.Main resultsThere were no RCTs; 8 high quality non-RCTs were included in the review. Meta-analysis showed reduction in mortality [N = 10109; OR = 0.47(0.39–0.56), p<0.00001; LOE: Moderate] and severe IVH and PVL [N = 5084; OR = 0.71(0.61–0.82), p<0.00001; LOE: Low] after exposure to ANC in neonates born <25 weeks. There was no significant difference in CLD [N = 4649; OR = 1.19(0.85–1.65) p = 0.31; LOE: Low] and NEC [N = 5403; OR = 0.95 (0.76–1.19) p = 0.65; LOE: Low]. Mortality was comparable in neonates born at 22, 23 or 24 weeks.ConclusionModerate to low quality evidence indicates that exposure to ANC is associated with reduction in mortality and IVH/or PVL in neonates born before 25 weeks.