Do drinkers respond to prices signals in the usual way by economising on beverages with higher prices and vice versa? Is the currency unit used in different countries irrelevant, or are drinkers subject to money illusion? Are the substitution effects of price changes symmetric? More fundamentally, can drinking patterns be adequately accounted for by the conventional utility-maximising approach? If so, how does consumption of beer, wine and spirits interact (if at all) in generating utility? According to the most recent data from the International Comparisons Program, on average, consumers in countries in the bottom quartile of the global income distribution devote something approaching one-half of all expenditures to food, while this falls to about 11% in the richest countries, in accordance with Engel’s law. The share for alcohol also drops, but much slower, so drinking rises noticeably relative to food as income increases. The within-alcohol distribution of spending (beer, wine and spirits) also changes quite dramatically. We use these cross-country data to address the above research questions. To visualise the data, we plot the budget shares for beer, wine and spirits in the form of a “drinking triangle”, which highlights the dominant beverage in each country. We also employ a Divisia-index-number approach to summarise the degree of price-quantity covariation. We cross-classify consumption and prices of beer, wine and spirits for nonparametric tests of the law of demand that higher prices lead to reduced consumption and vice versa. A system-wide model is estimated for the demand for beer, wine and spirits. As there is no unique way of ordering countries, a “levels version” of a differential system (similar to the Rotterdam model) is used. The system is used to test the hypotheses of homogeneity (the absence of money illusion) and symmetry of the substitution effects. For a substantial majority of countries, the Divisia price-quantity correlation is negative, which is suggestive evidence in favour of the economic approach to drinking. In the main, the results here support the law of demand. The hypotheses of homogeneity and symmetry cannot be rejected. Tests also reveal the coefficients are reasonable stable across countries, which sheds some light on the question of the similarity of tastes. Estimated price elasticities are tabulated for each beverage.