Of a total of 581 consecutive depressed subjects attending psychiatric services of 6 centres in Finland, 183 patients were eligible and completed a 6-week randomised double-blind trial with either moclobemide or fluoxetine. Of these, 32 (17.5%) patients met the DSM-III-R criteria for mood disorder with a seasonal (winter) pattern and 19 patients (10.5%) met the original criteria for seasonal affective disorder (SAD). There were no significant difference in the antidepressive response to the treatment between the patients with SAD and other depressive disorder. The treatment of 6 weeks with either moclobemide (300-450 mg daily) or fluoxetine (20-40 mg daily) resulted in a full remission in 15 (52%) patients with SAD and in 44 (37%) patients with other depressive disorder. Altogether, 23 (79%) and 83 (70%) patients, respectively, got significant benefit from the treatment. The improvement in the health-related quality of life in terms of ability to work was significantly more extensive in the patients with SAD assigned to receive moclobemide compared with the other depressive patients allocated to the same medication. Subjects who in addition met the operational criteria for atypical depression were equally distributed into the SAD and other depressive patient groups.