Objectives To identify the probability of survival and severe neurodevelopmental impairment (sNDI) at which perinatal physicians would or would not offer or recommend resuscitation at birth for extremely preterm infants. Methods A Delphi process consisting of five rounds was implemented to seek consensus (>80% agreement) amongst British Columbia perinatal physicians. The first-round consisted of neonatal and maternal-fetal-medicine Focus Groups. Rounds two to five surveyed perinatal physicians, building upon previous rounds. Draft guidelines were developed and agreement sought. Results Based on 401 responses across all rounds, consensus was obtained that resuscitation should not be offered if survival probability <5%, not recommended if survival probability 5 to <10%, resuscitation recommended if survival without sNDI probability >70 to 90% and resuscitation standard care if survival without sNDI >90%. Conclusions This physician consensus-based, objective framework for the management of an anticipated extremely preterm infant is a transparent alternative to existing guidelines, minimizing gestational-ageism and allowing for individualized management utilizing up-to-date data. Further input from other key stakeholders will be required prior to guideline implementation.