Background: Little is known about the influence of preference for a given therapy or preference against a modality of treatment on the outcome of that treatment. Results so far have been conflicting. The primary aim of this study was to investigate possible differences in outcome between panic disorder patients treated with preferred cognitive therapy and patients treated by randomization with the same intervention. Methods: A 12-week study comparing 35 patients treated by allocation with 31 patients treated by preference. Results: There were no differences at pretest between the two conditions on demographic or outcome measures. Outcome was assessed with measurements rating the panic frequency and severity of agoraphobia, general anxiety and depression. Both conditions improved significantly on nearly all ratings. There were no significant differences demonstrable between the two conditions on any of the outcome measures. Conclusion: Preference for a given therapy is not a powerful moderator of effect in the psychological treatment of panic disorder. Copyright (C) 2000 S. Karger AG, Basel.