The effect of self-discrepancy magnitude and salience on alcohol consumption was examined in an ad lib drinking study in order to evaluate the utility of the self-inflation component of the myopia model for better understanding drinking practices. Participants were 33 males and 27 females recruited on a university campus. It was predicted that participants with relatively large real self/ideal self discrepancies on dimensions important to their self-concept would consume the greatest amount of alcohol in a wine tasting test. Moreover, this effect was expected to be enhanced when self-discrepancies were made salient. The results of hierarchical regression analyses showed a main effect of gender and a significant interaction between self-discrepancy magnitude and salience condition. However, the interaction was such that wine consumption tended to decrease as discrepancy magnitude increased in the condition in which self-discrepancies were made salient, with the opposite relationship in the control condition. Three possible reasons for the unexpected findings are discussed: (a) The salience manipulation did not perform as expected: (b) the sample had little to gain from self-inflation: and (c) typically, self-inflation does not significantly motivate alcohol consumption.