Anxiety sensitivity (AS), the tendency to interpret feelings of anxiety as dangerous, is a core dispositional trait in a well articulated and extensively studied cognitive model of proneness to anxiety disorder. In recent years, there has been an increasing body of findings that also links AS to the tendency to use alcohol in general and the tendency to use alcohol as a means of coping with negative affect in particular. We expand on this empirical base by proposing and testing a theoretical model in which anxiety symptoms mediate the association between AS and alcohol use. That is, we propose that AS promotes anxiety symptoms, which, in turn, promote alcohol use aimed at coping with anxiety and other negative affect states. Over a 1-year data collection period, we assessed 82 alcohol-dependent individuals shortly after they began an intensive alcoholism treatment program. Self-reported anxiety symptoms associated with distinct anxiety syndromes were obtained with reference to the month period preceding their entry into the treatment program. Other information, including the presence of withdrawal symptoms, was obtained via interview. We found that syndrome-related anxiety symptoms and Trait Anxiety, but not State Anxiety or withdrawal symptoms, mediated the significant association between AS and the self-reported tendency to use alcohol as a means of controlling anxiety symptoms. Demonstrating a similar pattern of findings, but much less robustly so, were tests of these mediator models using alcohol use aimed at coping with negative affect (vs. coping with anxiety per se) as an outcome. In discussing these findings, we attempt to further develop a coherent model that incorporates AS, anxiety symptoms, and drinking motives. Our findings suggest that these relationships may differ for negative affect not specifically related to anxiety. We also discuss the possible associations of AS to withdrawal symptoms implied by our findings.