Using a controlled, clinical-trial format, 44 nonpsychotic, nonbipolar, depressed outpatients were treated with cognitive therapy or imipramine hydrochloride over a 12-week period. Although both interventions were associated with significant reductions in levels of depression, the cognitive-therapy patients showed greater symptomatic improvement and a higher treatment-completion rate. A one-year naturalistic follow-up of the 35 subjects who completed the protocol revealed that although many of the patients had a variable clinical course, both original treatment groups remained generally well. Self-rated depressive symptomatology was significantly lower for those who, one year earlier, had completed cognitive therapy than for those who had been in the clinical trial's pharmacotherapy cell. While there were several other interesting trends in favor of the cognitive-therapy patients, none of the between-group differences were significant. The pragmatic and clinical implications of the followup results are discussed.