In this paper, the demand for beer, wine, spirits and soft drinks in Ontario is modeled in two parts: an equation is specified to endogenize group expenditures and a demand system is set up to allocate budgeted group expenditures across ,types of beverages. Advertising is allowed to influence both the level of group expenditures and its allocation. Three popular advertising specifications are compared using the ]-test and the likelihood dominance criterion. Even though all three specifications fitted well according to standard criteria, the calculated expenditure, price and advertising elasticities were sensitive to the manner with which advertising is specified. This clearly highlights the need to rely on a sound criterion to identify a dominant specification. From the identified dominant specification, we found that advertising has very subtle effects on expenditures on alcoholic beverages (group and individual beverages). Thus, advertising is not effective in enlarging markets and this suggests that firms (especially breweries) use advertising to compete in zero-sum market share games. From a public policy perspective, our results are comforting but future research should investigate whether the neutral effect of advertising on aggregated expenditures hide substantial offsetting changes in the drinking habits of individuals.