Background: Transthoracic cardioversion fails to restore sinus rhythm in 6% to 33% of patients with atrial fibrillation. This study sought to determine the relative efficacy of biphasic waveforms compared with monophasic waveforms in the treatment of atrial arrhythmias. Methods: A total of 912 patients underwent 1022 transthoracic cardioversions between May 2000 and December 2001. A monophasic damped sine waveform was used in the first 304 cases, and a rectilinear biphasic defibrillator was used in the next 718 cases. Results: Use of a biphasic waveform was associated with 94% success in conversion to sinus rhythm compared with 84% with a monophasic waveform (P <. 001). The cumulative energy required to restore sinus rhythm was lower with biphasic shocks in both atrial fibrillation and atrial flutter groups (554 ± 413 J for monophasic vs 199 ± 216 J for biphasic shocks in the atrial fibrillation group, P <. 001; 251 ± 302 J vs 108 ± 184 J, respectively, in the atrial flutter group, P <. 001). In a multivariate analysis, use of a biphasic shock was associated with a 3.9-fold increase in success of cardioversion. Conclusion: When used to cardiovert atrial arrhythmias, the rectilinear biphasic waveform was associated with higher success rates and lower cumulative energies than the monophasic damped sine waveform.