Infants with intrauterine varicella-zoster virus (VZV) infection can suffer intracerebral VZV reactivations that require antiviral treatment.We report a 4-month-old boy whose mother had clinically confirmed mild varicella at 17 weeks' gestation, acquired from her first infant who had typical childhood chickenpox at the age of 5 years. At the age of 3 months, the infant developed typical dermatomal zoster rash within the ophthalmic division of the trigeminal nerve. On admission to hospital 1 day after onset of symptoms, the boy demonstrated right-accentuated clonic attacks. The EEG revealed delta waves with high amplitude and occasionally inserted spikes over the left occipito-parietal and temporal regions. MRI showed a signal alteration in the left region of the putamen interpreted as a post-ischaemic lesion (Fig. 1). Additionally, the ventricles were slightly enlarged. For laboratory studies, CSF and blood samples were taken 1 day after seizure onset. In the CSF, the number of cells (26 cells/ll) showing a lymphocytic cell picture was slightly elevated. Using the polymerase chain reaction, VZV DNA could be identified in the CSF. In a serum sample, specific IgM was negative and specific IgG concentration was elevated (enzyme-linked immunosorbent assay; Virotech, Ru¨sselsheim, Germany), indicating a past VZV infection. In the CSF, VZV-specific IgG class antibodies could be measured in titres of 1:64 with an indirect fluorescence antibody test. The boy was treated with phenobarbital, primarily 10 mg/kg as a single dose, followed by 5 mg/kg er day divided in two doses. For antiviral treatment, acyclovir was given intravenously at a dose of 10 mg/kg every 8 h for 10 days. Afterwards, the treatment with acyclovir was continued orally as suppressive therapy at a reduced dose for a period of 3 weeks. The drug was well tolerated and no signs of toxicity, in particular no signs of neutropenia, were seen. At the age of 6 months, VZV DNA could not be detected in the CSF and the titre of intrathecally measured VZV-specific IgG class antibodies decreased from 1:64 to 1:16. EEG findings did improve markedly at the age of 7 months and seizures did not occur.According to previous recommendations [4], diagnosis of congenital varicella syndrome (CVS) in our case was based on (1) the positive history of maternal varicella infection, (2) the presence of neurological defects, and (3) the proof of intrauterine VZV infection by the appearance of zoster during early infancy. Normally, CVS has been considered a multi-systemic disorder characterised by skin lesions accompanied with neurological defects, eye disease and/or limb hypoplasia [4]. In our case, the disease presented as generalised clonic seizures and the cerebral damage could be verified by both EEG and MRI findings. To date, manifestations of CVS limited to the brain have rarely been described in the literature [1,5,6]. In none of these reported cases was intracerebral infection proven by the detection of viral DNA and/or antibodies in the CSF. We could verify an intracerebral...