1. Twenty-two patients who met DSM-III-R criteria for panic disorder received either cholecystokinin 30-33 (25 micrograms i.v.) or 33% carbon dioxide. 2. The principal outcome measures of the study included the number and sum intensity of panic symptoms and the incidence of panic attacks. 3. The incidence of panic attacks tended (P = .07) to be higher with cholecystokinin 30-33 than with carbon dioxide. Nevertheless, patients who panicked within each group did not differ significantly with regard to the number and sum intensity of symptoms or symptom profile. 4. That cholecystokinin 30-33 and 35% carbon dioxide induced panic attacks which were qualitatively and quantitatively similar suggest that these agents might act on distinct systems that have a final common target or a final common mechanism of action.