This paper looks into the impacts of education and other household factors on alcohol consumption in Canada from 2004 to 2009. In terms of methodology, two-staged regression and quantile analysis are applied to an Almost Ideal Demand System (AIDS) that analyzes expenditure shares of alcoholic beverages. The Double Hurdle approach allows us to distinguish between the decision to consume (or not to consume) and the amount to be consumed. This is then complemented by an unconditional quantile regression analysis that permits us to differentiate between “heavy” and “moderate” consumers in a more general environment without results being conditional on specific population segments. The main findings include the importance of higher education for higher expenditure sensitiveness and lower expenditure shares, as well as higher consumption of alcohol and reduced sensitiveness to income for trade diploma graduates.