Oversupply has posed a number of problems for the Australian wine industry in recent times. When disaggregated from the industry level, however, the problem can be better described as a range of attribute-specific disequilibria. To date, solutions to this problem have predominantly revolved around reducing output through crop thinning or vine pulling. This paper proposes a different approach by suggesting that disequilibria may be reduced by gaining a better understanding of the demand for Australian wine. A discrete choice model of product differentiation is used to estimate the demand for wine in the United States, Australia's second largest export market. Implications of the analysis are explored.