We investigated the effects of a class I antiarrhythmic drug, cibenzoline, on human atrial muscle in vivo. Electrophysiologic measurements were performed in 44 patients (mean age 49 +/- 15 years), before and after an intravenous infusion of cibenzoline 1.4 mg/kg in 5 min. Extrastimuli at a basic cycle length (BCL) of 500 ms were delivered from the right atrial appendage. The effective refractory period of the right atrium (ERP-A), the conduction time from the high right atrium to the coronary sinus, maximum conduction delay (Max. CD), repetitive atrial firing zone (RAFZ), fragmented atrial activity zone (FAAZ), and conduction delay zone (CDZ) were measured. Patients were divided into two groups according to whether repetitive atrial firing (RAF) was induced (group A, n = 18) or not (group B, n = 26). Cibenzoline increased ERP-A from 198 +/- 25 to 214 +/- 26 ms (p < 0.05) and decreased Max. CD from 55 +/- 23 to 43 +/- 19 ms (p < 0.05). There were significant decreases in the RAFZ (10 +/- 17 to 4 +/- 10 ms, p < 0.05), the FAAZ (20 +/- 25 to 12 +/- 18, ms p < 0.05), and the CDZ (41 +/- 21 to 32 +/- 19 ms, p < 0.05). Cibenzoline significantly increased ERP.A (186 +/- 25 to 212 +/- 26 ms, p < 0.05) in group A, but not in group B. There were significant decreases in the RAFZ [25 +/- 19 to 9 +/- 15 ms (p < 0.05) and FAAZ 22 +/- 29 to 11 +/- 21 ms, (p < 0.05)] in group A, but not in group B. The results suggest that cibenzoline can suppress paroxysmal atrial fibrillation by prolongation of ERP-A and may also have preferential effects on the substrate of atrial fibrillation and RAF.