Background: In preterm infants with premature rupture of membranes, antibiotic treatment is frequently started but rates of early onset sepsis are lower. In line with national guidelines, a stratified approach in the decision to start antibiotic treatment using maternal history, clinical impression and biomarkers has been implemented in our level III neonatal center and its results are evaluated. Methods: Retrospective cohort study of all preterm newborns with rupture of membranes at least 1 h prior to delivery admitted to our tertiary neonatal intensive care unit. Data on antibiotic exposure, mortality and major neonatal complications were extracted from the electronic patient charts to evaluate the effects and safety of our stratified approach. Results: Four hundred fifty-six infants met the inclusion criteria. 120 (26%) received primary antibiotics whereas 336 (74%) did not. Of those receiving primary antibiotics, 13 (11%) had a blood culture positive sepsis, 46 (38%) met the criteria of clinical sepsis and in 61 (51%) sepsis was ruled out and antibiotics were stopped after 48-96 h. All infants with blood culture positive sepsis were identified and treated within the first 24 h of life using this approach. None of the 336 infants who were not started on antibiotics primarily needed antibiotic therapy within the first 5 days of life. There were no deaths or major neonatal complications in the group that did not receive empiric antibiotics. Conclusions: Our stratified approach for preterm infants with premature rupture of membranes allows a safe reduction of antibiotic exposure even in this high risk population. As a result, only 25% of high risk preterm newborns are treated with antibiotics of which more than half receive less than 5 days of treatment. To treat one infant with blood culture positive sepsis, only 9 infants receive empiric antibiotics.