Very low-birthweight infants constitute more than one-quarter of all new cases of cerebral palsy. We performed a case-control study of associations between antenatal maternal infection and cerebral palsy in very low-birthweight infants. Cases and controls were selected from a cohort of 1238 consecutive infants who: (1) had birthweights between 500 and 1500 g and no major congenital anomaly; (2) were born 1 January 1986 to 31 December 1993 to a mother residing in 1 of 17 counties in north-west North Carolina; and (3) were delivered at the only tertiary obstetric referral centre in those same 17 counties. A total of 984 of these infants (79%) survived to 1 year of age (adjusted for degree of prematurity) and were scheduled for a multidisciplinary examination; 815 (83%) came as scheduled. Excluding two cases attributable to post-neonatal events, 62 cases of cerebral palsy were identified. Controls were the two infants, without cerebral palsy, born closest in time to each case. Medical records were reviewed by a nurse who was not aware of which subjects were cases. Among possible markers of intra-amniotic infection, those associated most strongly with cerebral palsy were chorioamnionitis diagnosed by an obstetrician (odds ratio [OR] adjusted for gestational age [95% confidence limits] = 2.6 [1.0, 6.5]), antepartum maternal temperature > 37.8 degrees C (OR = 2.6 [1.1, 6.0]), uterine tenderness (OR = 2.6 [0.8, 9.3]), maternal receipt of antibiotics (OR = 2.2 [1.0, 4.7]) and neonatal sepsis in the first week of life (OR = 2.9 [0.9, 8.9]). All of these associations were stronger for diplegia than the other clinical subtypes of cerebral palsy. The association with chorioamnionitis and spastic diplegia persisted when adjusted for maternal magnesium sulphate receipt, maternal betamethasone receipt, method of delivery (vaginal vs. abdominal), acidosis on the neonate's initial arterial blood gas, systolic blood pressure < 30 mmHg and the diagnosis of major neonatal neurosonographic abnormality.