It is unknown whether bepridil improves cardiovascular events in atrial fibrillation (AF) patients, so this study evaluated the clinical outcome in paroxysmal or persistent AF patients receiving bepridil. We conducted a cohort study of 284 consecutive patients who received bepridil for AF (25% female, 5913 years) with a median follow-up period of 17 months (4-157 months). A total of 135 (48%) patients had structural heart disease, and 231 patients (81%) had previously received class I or class III antiarrhythmic drugs. The cumulative rates for cardiovascular events were 2.4%, 8.1%, and 10.1% at 1, 3, and 5 years, respectively. The cumulative rates for a composite of mortality, cerebral infarction, systemic embolism, major bleeding and heart failure were 9.7%, 18.2%, and 29.6% at 1, 3, and 5 years, respectively. The probability of progression to permanent AF was 23.5% at 5 years. Sudden death occurred in a patient with a prior myocardial infarction who was taking 200mg daily, and torsade de pointes (Tdp) occurred in two patients without structural heart disease taking 200mg daily. Excessive corrected QT interval prolongation (>0.50s) was observed when plasma concentrations were higher than 800 ng/ml. Bepridil might not improve the clinical outcome in refractory AF patients. Bepridil-related adverse events, including QT prolongation and Tdp, occurred in a dose- and concentration-dependent manner.