This dissertation deals with the impact of alcohol advertising, which associates drinking alcohol with many desires and values psychologically appealing to young people. As traditional research on alcohol advertising's impact focused on a direct link between exposure and alcohol consumption, this study breaks the link into several mediating stages. Under the theoretical framework of the Message Interpretation Process (MIP) model, this study involves how college students interpret alcohol advertising, how their desire, wishful thinking and expectancies toward drinking may account for both the impact of exposure to alcohol advertising and reasons for drinking. Specifically, the purpose of the study is to (1) assess the MIP model and (2) explore the impact of anti-alcohol messages on college students' interpretation of alcohol advertising. An experiment was conducted with 94 college students. In the experiment, participants were grouped into three treatment groups: the first was exposed to logic-based anti-alcohol messages followed by alcohol advertisements, the second was exposed to emotion-based anti-alcohol messages followed by alcohol advertisements; and the third group was exposed to alcohol advertising only, as the control group. Results showed support for the MIP model in general. As posited by the model, realism and desirability are both related to identification with characters seen in alcohol advertising, which in turn is related to expectancies toward drinking, which is in turn related to intention to drink. The only two exceptions were the lack of significance for the social norm-identification link and relationships with negative expectancies. Compared with exposure, realism and desirability were better predictors for alcohol advertising's impact. The results also showed impact of logic-based anti-alcohol messages on identification and expectancies as well as impact of emotion-based anti-alcohol messages on desirability and identification. There was no significant difference between the two anti-alcohol groups. This study provides insights on how viewers draw from alcohol advertising to make sense of their own lives. Implications for anti-alcohol campaign are provided. Because negative expectancies toward drinking did not influence the intention to drink, the association between negative outcomes and drinking might not very well. Another association between positive outcomes and non-drinking may work better. Recommendations for future research as well as limitations of this study are discussed.