INTRODUCTION AND AIMS: To examine heavy episodic drinking across demographic subgroups to identify where heavy episodic drinking is socially located in an Australian state. DESIGN AND METHODS: Cross-sectional survey, 2483 adult Victorians using Computer Assisted Telephone Interviewing. Two measures of heavy drinking were used: (i) heavy episodic drinking-more than five standard drinks at least weekly; and (ii) typically heavy drinking-50% or more of all drinking occasions involving consumption of 5+ standard drinks. Associations between heavy episodic drinking and eight potential sociodemographic correlates (gender, age, education, income, marital status, ethnic origin, religion and geographical remoteness) were examined. RESULTS: There were few significant correlates of heavy episodic drinking apart from gender and age, once gender and age were controlled. Men were more likely to report heavy episodic drinking than women, but this was attenuated in the measure of typically heavy drinking, suggesting that women reporting heavy episodic drinking were more likely to typically drink that much when they drank. Younger people were more likely to report weekly heavy episodic drinking and more likely to report engaging in this pattern on at least half of their drinking occasions, and this was also true for those unmarried or in de facto relationships. Those of Asian background were less likely to report heavy drinking. In multivariate analysis, the remaining sociodemographic variables were largely unrelated to the drinking measures. DISCUSSION AND CONCLUSION: The study clearly shows that the prevalence of heavy episodic drinking varies particularly across gender and age groups in Victoria. These variations appear to hold across key sociodemographic variables such as income and education.