To investigate the effect of verapamil on atrial vulnerability, the following measurements were performed before and after the intravenous administration of verapamil (0.15 mg/kg) in 10 subjects with paroxysmal atrial fibrillation (Paf), and 10 subjects without Paf (non-Paf). During the sinus rhythm, 1) intra-/interatrial conduction time (Intra-/Inter- ACT); the initial deflection of high right atrium (HRA) to that of His bundel/coronary sinus were measured. After 8 consecutive HRA stimuli (A1), premature stimulus (A2) was introduced by shortening the coupling interval (A1A2) and we measured 2) conduction delay zone (CDZ); the zone of A1A2 with the prolongation of Inter-ACT, 3) % maximum atrial fragmentation (%MAF); % maximum value of the ratio of HRA activity width at A2 (Awt) against that at A1 (Awc), 4) fragmented atrial activity zone (FAZ); the zone of A1A2 with % value of Awt/Awc more than 150%,5) repetitive atrial response (RAR); more than 2 atrial activities which occur in response to A2. (6) right atrial effective refractory period (RAERP). Verapamil significantly shortened CDZ and %MFA, slightly lengthened RAERP, and had RAR disappear in Paf, while it did not effect any indices significantly in non-Paf. We concluded that verapamil could reduce atrial vulnerability in Paf due to blocking of atrial conduction delay mediated by slow response fibers.