AIMS: With transthoracic cardioversion of atrial fibrillation (AF), biphasic are more effective than monophasic waveforms. We sought to determine the ideal energy levels for biphasic waveforms. Methods We compared biphasic truncated exponential waveforms with monophasic damped sine waveform defibrillators, in a prospective, single-centre, randomized (1:1 ratio) study. The study included 154 patients receiving concomitant amiodarone; 77 received serial biphasic (50, 100, 150, up to 175 J) and 77 monophasic shocks (100, 200, 300, up to 360 J), as necessary. Results First-shock efficacy was similar in the two groups (57 vs. 55%, P = 0.871, respectively), as were serial-shocks (90 vs. 92%, P = 0.780). Both groups received equal numbers of shocks (1.8 +/- 1.1 vs. 1.7 +/- 1.0, P = 0.921). In both groups, serum creatine kinase levels showed a small but significant increase. The increase was, however, higher in the monophasic group. CONCLUSION: In patients with concomitant amiodarone therapy, biphasic truncated exponential shocks, using half the energy, were as effective as monophasic damped sine shocks. The biphasic scheme was not more efficacious for cardioverting AF. In our population, a first shock of at least 100 J seemed advisable with either waveform. If necessary, escalating shocks must be performed, but ideal levels of increase per shock are still uncertain for biphasic waveforms.