OBJECTIVE: To establish the prevalence of fetal alcohol syndrome (FAS) in Victoria through the Victorian Birth Defects Register (VBDR). METHODS: A sample of live births from 1995-2002 was selected from the Victorian Perinatal Data Collection and VBDR based on reported microcephaly, FAS or maternal use of alcohol during pregnancy. Following ethics approval, medical records of mother and child were requested for 117 births. One hundred and nine of these were accessed and examined for factors related to FAS. Records were categorised as FAS, possible FAS, unable to categorise, or not FAS. RESULTS: From the VBDR the prevalence was calculated at 0.006 per 1,000 live births. Four additional possible cases of FAS increased this to 0.014 per 1,000 live births. Six cases were defined as 'unable to categorise' as alcohol use was unknown but other features of FAS were evident. Including these cases, plus five where some low-level alcohol use was reported, increased the prevalence to 0.03 per 1,000 live births. Twenty-eight per cent of the audit population and 39% of the microcephalic cases had no information about maternal alcohol use recorded in the antenatal or babies' records. CONCLUSION AND IMPLICATIONS: The audit of medical records provided additional information regarding FAS prevalence in Victoria. This prevalence ranges from 0.01 to 0.03 per 1,000 live births. To accurately assess the extent of the problem, there needs to be improved reporting of alcohol use in pregnancy and a system in place to report cases diagnosed during and beyond the perinatal period to the VBDR.