The "right" tax is difficult to estimate precisely because reliable data are lacking on the drinking patterns and responsiveness of adults, and there is no clear understanding of the associations between drinking and driving for all age groups. Consumers' own perceptions of risk enter the calculation of the proper tax level but also introduce significant alternative policy options to taxation. If consumers systematically underestimate the risk, then investments in education may prove fruitful. Comments accompanying the paper pertain to studies on the relationship between alcohol prices and consumption demand, the need for better data about the relative price sensitivity of alcohol for various demographic groups, and the need for appropriate preventive programs to accompany an alcohol tax. 34 references, 30 notes, 2 tables.