This study examined the effects of alcohol outcome expectancy on panic symptoms, subjective anxiety, and heart rate after a 35% carbon dioxide panic challenge in patients with panic disorder after consuming either an alcoholic or placebo beverage. Competing predictors of challenge response included beverage group and anxiety sensitivity. We found that alcohol outcome expectancies predicted severity of self-reported panic symptoms after the challenge, while controlling for alcohol and anxiety sensitivity; participants with greater positive expectancies reported fewer panic symptoms. In addition, participants who drank alcohol reported less subjective anxiety after the challenge, and self-reported anxiety prior to the challenge mediated this relationship; individuals with greater anticipatory anxiety reported greater anxiety after the challenge. Anxiety sensitivity did not predict challenge response. The results suggest that individual differences in alcohol outcome expectancies, as well as pharmacological effects of alcohol use, influence the experience of panic in patients with panic disorder.