ObjectiveTo determine if postnatal transfer or birth in a non-tertiary hospital is associated with adverse outcomes.DesignObservational cohort study with propensity score matching.SettingNational health service neonatal care in England; population data held in the National Neonatal Research Database.ParticipantsExtremely preterm infants born at less than 28 gestational weeks between 2008 and 2015 (n=17 577) grouped based on birth hospital and transfer within 48 hours of birth: upward transfer (non-tertiary to tertiary hospital, n=2158), non-tertiary care (born in non-tertiary hospital; not transferred, n=2668), and controls (born in tertiary hospital; not transferred, n=10 866). Infants were matched on propensity scores and predefined background variables to form subgroups with near identical distributions of confounders. Infants transferred between tertiary hospitals (horizontal transfer) were separately matched to controls in a 1:5 ratio.Main outcome measuresDeath, severe brain injury, and survival without severe brain injury.Results2181 infants, 727 from each group (upward transfer, non-tertiary care, and control) were well matched. Compared with controls, infants in the upward transfer group had no significant difference in the odds of death before discharge (odds ratio 1.22, 95% confidence interval 0.92 to 1.61) but significantly higher odds of severe brain injury (2.32, 1.78 to 3.06; number needed to treat (NNT) 8) and significantly lower odds of survival without severe brain injury (0.60, 0.47 to 0.76; NNT 9). Compared with controls, infants in the non-tertiary care group had significantly higher odds of death (1.34, 1.02 to 1.77; NNT 20) but no significant difference in the odds of severe brain injury (0.95, 0.70 to 1.30) or survival without severe brain injury (0.82, 0.64 to 1.05). Compared with infants in the upward transfer group, infants in the non-tertiary care group had no significant difference in death before discharge (1.10, 0.84 to 1.44) but significantly lower odds of severe brain injury (0.41, 0.31 to 0.53; NNT 8) and significantly higher odds of survival without severe brain injury (1.37, 1.09 to 1.73; NNT 14). No significant differences were found in outcomes between the horizontal transfer group (n=305) and controls (n=1525).ConclusionsIn extremely preterm infants, birth in a non-tertiary hospital and transfer within 48 hours are associated with poor outcomes when compared with birth in a tertiary setting. We recommend perinatal services promote pathways that facilitate delivery of extremely preterm infants in tertiary hospitals in preference to postnatal transfer.