In a prospective randomized study we searched for arrhythmogenic effects of the tetracyclic antidepressant, a maprotiline, and the tetrahydroisoquinoline derivative, nomifensine. Forty depressive patients from the psychiatric outpatients department were included in the study. Twenty patients in each group received maprotiline or nomifensine over three weeks in the recommended daily dosage of 75 mg. Rhythm analysis was performed before therapy, at the end of 3 weeks therapy, and 1 week after withdrawal from medication using a dual channel long-term ECG with monitoring periods of 10 h during normal daily activities. Before treatment, spontaneous incidence of all ventricular ectopics and of their complex forms was within the normal range when compared with ectopic activity of 121 "normal subjects" without detectable heart disease. No significant increase could be demonstrated during therapy with maprotiline or nomifensine, nor was any change observed 1 week after medication had been stopped. The same was true for supraventricular extrasystoles; atrial tachycardia, atrial flutter, and fibrillation were never seen. Sinoatrial (n=2) and atrioventricular block (n=1) were rare findings independent of and not affected by treatment. No bundle branch blocks were observed before, during, and after treatment. In contrast, despite the conservative dosage of both drugs, a therapy-dependent increase in average heart rate was found (p less than 0.001). This increase was significantly higher in patients receiving nomifensine than in those treated with maprotiline (p less than 0.001), suggesting a lower intrinsic anticholinergic activity of the latter compound.