During a 3 year period, seventy patients aged 53 +/- 16 years with a total of 73 arrhythmias were treated over a mean period of 6.8 months (maximum 27 months) with oral propafenone, the usual dose being 900 mg/day. The study covered the whole spectrum of cardiac arrhythmias (32 supraventricular, 41 ventricular), and their relation to the autonomic nervous system. The efficacy was scored from 1 (no effect) to 5 (complete control) as judged by the clinical response, the results of Holter monitoring (175 control and 133 test recordings on therapy), and a comparison was made between the effects of propafenone and other antiarrhythmics: quinidine-like drugs, beta-blockers and amiodarone. With respect to supraventricular arrhythmias: 9 cases of vagally-induced atrial flutter and fibrillation were unaffected by propafenone (mean score = 1.1). On the other hand, the drug was very effective (mean score = 4.1) in 8 cases of adrenergic atrial arrhythmias. In 12 arrhythmias with more varied mechanisms (extrasystoles, tachysystole, paroxysmal atrial fibrillation) an intermediate score was obtained (2.8). Three cases of resistant junctional tachycardia due to reentry were improved. At ventricular level, 5 cases of extrasystole sensitive to quinidine were also improved by propafenone (4.6); the difference was more clearcut in 8 cases of benign idiopathic tachycardia (propafenone: 4.1, and quinidine: 2.4). This was more marked in 13 cases of more severe arrhythmia in diseased hearts in which the effect of propafenone (4.1) was superior even to that of amiodarone. However, propafenone was less effective (3.3) than amiodarone in 4 cases of severe polymorphic idiopathic ventricular tachycardia closely related to the autonomic nervous system. The antiarrhythmic effect of propafenone was appreciable in 10 cases of resistant post-infarction ventricular tachycardia, eventually in association with amiodarone. Slowing of the sinus rhythm (-11.6%) with no change in the day/night ratio was due to beta-inhibition. However, in toxic doses this may progress to sinoatrial block (9 cases). A lengthening of the PR interval and duration of QRS was common, but this was not complicated by torsade de pointes, one case of which was successfully treated by propafenone. Secondary gastro-intestinal effects and vertigo were rarely severe enough to warrant stopping therapy. In conclusion, these results show that the introduction of propafenone is a valuable therapeutic advance in the treatment of arrhythmias, especially in those with a favoring adrenergic mechanism.