We assessed the tolerability and utility of switching fluoxetine nonresponders to citalopram the day that fluoxetine therapy was stopped. Fifty-eight outpatients with DSM-IV major depressive episode and prospectively confirmed nonresponse to fluoxetine (mean final dose = 31 mg/day) were switched directly to citalopram (20 mg/day). Of the 58 patients, 44 (76%) had never been successfully treated with antidepressant medication. During a 12-week open-label treatment period, citalopram could be titrated up to a maximum dose of 60 mg/day. Response was evaluated using the Clinical Global Impressions (CGI) scale, the 24-item Hamilton Rating Scale for Depression, and several other measures. Eighty-one percent (N = 47) completed the trial, and citalopram (mean dose = 38.8 mg/day) was well tolerated. The intent-to-treat CGI response rate was 46% (26/57) at week 6 and 63% (36/57) at study endpoint; the completer response rate was 76% among the 47 patients who completed the 12-week trial. Improvement from baseline on all dependent measures was statistically significant after the first week of citalopram treatment. Fluoxetine nonresponders can be quickly switched to citalopram, with good tolerability and reasonable chance of therapeutic benefit. Further work is necessary to assess the merits of this treatment strategy relative to other options.