Fifty-two patients with depressive illness characterized by four symptoms (periodical course, psychomotor retardation, diurnal variation, unrealistic self-depreciation) and a score of at least 18 on the Hamilton Depression Scale 1-17 (HDS) were allocated to a double-blind randomized study with femoxetine and imipramine. Patients were diagnosed according to RDC and further classified according to the Newcastle-II index. During the six weeks of treatment, efficacy was evaluated by means of HDS and a global evaluation. Side-effect symptoms were recorded on a check-list by questioning. After six weeks of treatment with femoxetine or imipramine (recommended daily standard dosages are 600 mg femoxetine and 150 mg imipramine (b.i.d.); in the present study, dosages were flexible and could be adjusted according to effect/side-effects) evaluation of efficacy based on HDS, a six-item subscale, groups of HDS items as well as single items showed no statistically significant differences between the treatment groups except with regard to the factor for sleep disturbances in the HDS, where greatest reduction was seen in the femoxetine group. No statistically significant differences regarding side-effect profile were seen. However, in the imipramine group, higher frequencies of such moderate to severe symptoms as dry mouth, constipation and urination difficulties were observed (the greatest difference was seen for dry mouth, p 0.1, while p-values for the remaining two symptoms were greater than 0.1). Moreover, based on the patients' own opinion on side-effects, femoxetine seemed to be better tolerated. One patient took an overdosage of approx. 26 g femoxetine; half of the intake was removed by gastric emptying at the hospital.(ABSTRACT TRUNCATED AT 250 WORDS)