The results of a caffeine consumption inventory indicated that patients with panic anxiety disorder, but not affectively ill patients or normal controls, had levels of self-rated anxiety and depression that correlated with their degree of caffeine consumption. In addition, this self-report survey suggested that patients with panic disorder had an increased sensitivity to the effects of one cup of coffee. This apparent sensitivity to caffeine was also documented by the observation that more patients with panic disorder reported the discontinuation of coffee intake due to untoward side effects than controls. These results, based on self-reports, suggest that the hypothesis that patients with panic disorder are more reactive to caffeine should be directly tested using caffeine challenges and that the mechanisms underlying caffeine's effects on anxiety should be further explored.