Aims. This paper explores the relationship between alcohol sales and assault in New South Wales (NSW), Australia. Design. The design was a cross-sectional analysis of the relationship between alcohol sales and assault, taking into account spatial autocorrelation. Setting. The analyses were carried out separately on data from metropolitan Sydney and country NSW. Participants. The unit of analysis was local government areas (178 in NSW). Measurements. The dependent variable was assault rates. The independent variables were overall alcohol sales and sales disaggregated by alcohol and outlet type and relevant demographic and socio-economic factors. Findings. Multiple regression analyses revealed a significant relationship between overall alcohol sales in an area and its incidence of assault for both Sydney and country NSW. There were few effects of outlet or alcohol type for Sydney, but the effect of alcohol sales from hotels and off-licences accounted for most of the alcohol-assault relationship for country NSW. Conclusions. Although the current research cannot exclude an opportunity account of the role of alcohol in assault, the results are entirely consistent with alcohol consumption as a causal agent.