The goal of the present research was to test the value of evaluative conditioning (EC) to unobtrusively change alcohol-related attitudes and drinking behavior. In the EC paradigm, participants had to spot an irrelevant target picture in a series of trials in which many different stimuli were presented. In the experimental condition, beer-related pictures (CSs) were consistently paired with negative words and pictures (USs) in a number of trials. In the control condition, participants were exposed to the same stimuli, but without the critical alcohol-negative pairings. After the EC task, participants participated in an allegedly second experiment in which we measured beer-related attitudes, craving for beer, and actual drinking behavior both during a bogus taste test and during the week following the experiment. Compared to participants in the control condition, participants in the experimental condition showed more negative attitudes toward beer, experienced less craving for beer, and consumed less beer both in the lab during the taste test and outside the lab during the week following the manipulation. These findings suggest that unhealthy drinking behavior may be targeted through EC procedures.