There is now substantial evidence to suggest that the presence of inflammatory molecules within the fetal circulation and brain is associated with damage to vulnerable oligodendrocyte precursor cells before 32 weeks' gestation and long-term white matter damage (WMD); however, the nature of the inflammatory stimulus remains unclear. The common association between ascending bacterial infection from the vagina, haematogenous spread of periodontal organisms, chorioamnionitis, preterm delivery, and neonatal brain injury provides indirect clinical evidence supporting an important role for bacteria. Animal experiments in which live bacteria, or endotoxins, lead to the classical focal necrotic lesions of periventricular leukomalacia add further support to this concept. The isolation of bacteria-specific DNA from only a small proportion of cases where there is evidence of histological chorioamnionitis, suggests that there may be non-bacterial causes of fetal inflammation. Viruses make attractive alternative candidates; they are a common cause of maternal pyrexia, can cross the placenta (e.g. cytomegalovirus and rubella), are known to cause WMD and microcephaly, and are difficult to isolate. A recent study of viral DNA in stored neonatal blood spot specimens suggested a possible link between viral exposure and later cerebral palsy, particularly in preterm babies. Dammann and Leviton have been even more specific, identifying maternal pestivirus (PV) infection as a hypothetical cause of WMD in preterm neonates.