OBJECTIVE: The aim of this study was to investigate to what extend perinatal factors contribute to the neurodevelopmental outcome in a group neonates born after spontaneous preterm labour with or without prolonged rupture of the membranes (PROM). METHODS: In a cohort of neonates born after the spontaneous onset of labour with or without PROM before 34 weeks of gestation a stepwise forward logistic regression was performed to analyse the influence of antenatal and postnatal variables on adverse outcome. Adverse neurodevelopmental outcome was defined as a Griffith's developmental score <85, cerebral palsy, a major disability or perinatal death associated with severe cerebral damage. RESULTS: The study group consisted of 185 neonates. Seven neonates died with severe cerebral damage. After a forward logistic regression analysis three factors appeared to have an independent influence: gestational age protected against an adverse outcome (odds ratio (OR) per day increase 0.95, 95\% confidence interval (CI) 0.90-0.97) while abnormal cranial ultrasound (intraventricular haemorrhage and periventricular leucomalacia) (OR 6.33, 95\% CI 2.16-18.52) and the need for a second course of antibiotics (OR 1.85, 95\% CI 1.02-3.33) increased the risk for adverse outcome. Comparing the group with a normal neurodevelopmental outcome with those with cerebral palsy, cranial ultrasound abnormalities were independently associated with cerebral palsy (OR 48.75, 95\% CI 11.78-201.76). CONCLUSION: The most important way of preventing neurological damage in infants is to increase gestational age at birth and to avoid the development of intraventricular haemorrhage and periventricular leucomalacia.