Background: Because of a possible risk of induction of Ventricular Fibrillation (VF) by defibrillation of atrial fibrillation (AF) postulated by LOWN and coworkers, synchronized cardioversion is used worldwide. This prospective, randomized study assessed the efficacy and safety between R-wave controlled cardioversion and defibrillation of AF at 2 study centers in Cologne, Germany. Hypothesis: Defibrillation is not significantly different from cardioversion primarily in the occurrence of VF or sustained Ventricular Tachycardia (VT) and secondarily in restoring sinus rhythm, inducing non-sustained VT, asystole, or bradycardia. Methods: 146 patients at an outpatient practice and 122 at the university hospital were randomized to cardioversion (n = 140) or defibrillation (n = 124). Results: Cardioversion was successful in 92.1% of cases and defibrillation in 87.1%. The difference in efficacy was not statistically significant. In n = 1 patients receiving defibrillation, VF occurred after the first shock (200J) and immediate defibrillation (200J) restored sinus rhythm. In the n = 1 case, asystole occurred during cardioversion which terminated spontaneously. In n = 1 patients cardioverted and n = 2 who were defibrillated, sinus bradycardia occurred requiring Atropine in two cases. There were no thromboembolic events within 10 days. N = 9 patients reverted to AF within two hours. No patients died. Conclusion: Electrical conversion of AF can be performed with similar results and low risk with both R-wave-triggered cardioversion and defibrillation. In particular, defibrillation with higher energies (> 100J) can be performed as effectively and safely without a statistically significant increased risk of VF or VT. There was no difference in efficacy and risk between electrotherapy performed in the outpatient and inpatient settings.