One of the major approaches to alcoholism prevention is referred to as the distribution of consumption model. This prevention model can be summarized as a causal model whereby the availability of alcoholic beverages has a direct causal effect on the aggregate level of alcohol consumption in the population and, in turn, an indirect effect on the incidence and prevalence of alcohol-related damage. This article summarizes an application of a statistical technique known as the analysis of linear structural relations (LISREL) to a set of Ontario data concerning alcohol availability, alcohol consumption and alcohol-related damage. Results substantiated the existence of specific causal paths consistent with the model. Several procedures for assessing the overall goodness-of-fit of the model suggested that it adequately fit the data. The results provide reasonable statistical evidence that government policies restricting the retail availability of alcoholic beverages will reduce the per capita rates of alcohol consumption and, in turn, reduce the level of alcohol-related mortality and morbidity in the general population.