With regard to the restrictive and safe use of class I-antiarrhythmics (AA) clinical criteria predicting drug efficacy and safety have become important. In 140 pts (102 men, 63 ± 7 years) with frequent and symptomatic, but not life-threatening arrhythmias (VA; > 30/h) occurrence, reproducibility and therapeutic consequences of a circadian distribution (CD) of single/repetitive VA was studied. Pts underwent 3 24 h-Holter recordings at baseline, after 2 weeks of placebo and after 6 weeks of dose-titrated antiarrhythmic therapy. 67/140 pts (48%) had reproducible day-night CD (daytime VA increase: 50/67 pts, 75%). In Cox's regression analysis CD was characteristic of a lower VA rate (p < 0.05), but more frequent tachycardia episodes (risk: 3.8 times, p < 0.05) and a lower ejection fraction. AA response was better in pts with CD (3.3 times, p < 005). Day-/night CD pattern was also predictive for a superior response of class IA or IC agents (Fig.). In conclusion CD of VA is common in half of the pts with symptomatic VA. When class I agents are used. CD provides clinical evidence for selection of the most effective AA and dose and may help to improve overall antiarrhythmic benefit.