In patients with malignant ventricular arrhythmias, antiarrhythmic therapy is known to carry a substantial risk of proarrhythmia. This risk is usually considered to be low when supraventricular arrhythmias or benign ventricular arrhythmias are considered. We were able to collect data on four patients without a history of life-threatening arrhythmias, in whom antiarrhythmic therapy was used and resulted in documented ventricular fibrillation or torsade de pointes. In Cases No. 1 and 2, atrial fibrillation was treated with either quinidine or quinidine and sotalol in combination. In both patients Holter monitoring, 4-12 h after conversion to sinus rhythm, documented the spontaneous occurrence of torsade de pointes degenerating into ventricular fibrillation and requiring DC shock for termination. In Case No. 3, atrial fibrillation was treated with sotalol and amiodarone for 2 months when incessant episodes of torsade de pointes were documented. In Case No. 4, frequent but unsustained ventricular arrhythmias were treated with amiodarone in a patient suffering dilative cardiomyopathy. After 6 days of treatment at a heart rate of 54 beats/min, a marked QT increase was associated with the occurrence of repetitive episodes of polymorphic ventricular tachycardia degenerating into ventricular fibrillation. None of the patients presented significant electrolyte abnormalities in the laboratory. A pathologic increase of the QTc-time was documented in Cases No. 1, 3, and 4. In all patients antiarrhythmic therapy was withdrawn after the proarrhythmic event and the patient became free of malignant tachyarrhythmias. Antiarrhythmic therapy also carries a considerable risk of proarrhythmia when "benign" cardiac arrhythmias are treated. The risk seems to be lower than in patients with malignant arrhythmias, however it includes the occurrence of lethal tachyarrhythmias. Special attention should be paid to the selection of antiarrhythmic agents when used in combination.