Sixty-two agoraphobic patient who had completed a controlled study of therapist-assisted in vivo exposure (flooding) and imipramine were assessed 1 month, 6 months, 1 year, and 2 years later. Overall, improvement during treatment was maintained throughout follow-up. At 1 month but not subsequently, imipramine and flooding had significant effects on central measures of agoraphobia. Patients who were marked treatment responders had a favorable clinical course and did not experience secondary depression, unlike patients who had not responded markedly to treatment. These findings suggest that treatments which evoke maximum therapeutic benefit initially are likely to foster long-term maintenance and reduce subsequent depressive sequelae.