Atrial fibrillation (AF) is an extremely common arrhythmia seen in clinical practice. Because of the limited efficacy of traditional therapeutic strategies to restore and maintain normal sinus rhythm, several nonpharmacologic options have evolved. The promising results achieved with internal atrial defibrillation have facilitated the development of an implantable atrial defibrillator. Preliminary results obtained from an initial study on a small number of highly selected patients with refractory AF suggest that atrial defibrillation can be performed effectively and safely with adequate patient tolerance by using a stand-alone device. The extension of this therapy will depend on the results of well-designed prospective studies comparing this new therapeutic option with traditional methods. Several acute studies have shown that internal conversion of AF is feasible at low energies with current endocardial transvenous lead configurations primarily designed for ventricular defibrillation, but long-term efficacy has, to date, only been demonstrated with atrial implantable defibrillator lead systems. As AF is a frequent arrhythmia in implantable cardioverter defibrillator (ICD) recipients, it would seem desirable to incorporate the capability for atrial defibrillation into an ICD. Clinical studies have shown that an atrial defibrillator, as part of a combined dual-chamber ICD system, may not require a potentially complicated switching network for establishing different electrode configurations for atrial and ventricular tachyarrhythmia. The efficacy in atrial cardioversion of such a combined, less complex device seems to be as high as reported for a pure atrial defibrillator, but generally at somewhat higher energy requirements. The results of further investigations will show whether a dual-chamber cardioverter defibrillator would be of clinical relevance in patients with ventricular and supraventricular tachyarrhythmia.