Direct-current electrical cardioversion is the main method for the conversion of atrial fibrillation. Its success depends on many factors. In several studies, biphasic shock waveforms have been demonstrated to be superior to monophasic shocks for termination of atrial fibrillation; however, information about impact of paddle position is controversial. Initial energy level is an object of discussions. The aim of the study was to compare a truncated exponential biphasic waveform with monophasic damped sine waveform and antero-lateral with antero-posterior paddle positions for cardioversion of atrial fibrillation, to determine its impact on early reinitiation of atrial fibrillation. A total of 224 consecutive patients with atrial fibrillation underwent electrical cardioversion with biphasic (Bi, n=112) or monophasic (Mo, n=112) shock waveform in a randomized fashion. The position of hand-held paddle electrodes was randomly selected in both groups to be anterior-lateral and anterior-posterior. Energies used were 100-150-200-300-360 J (Bi) or 100-200-300-360 J (Mo). If monophasic shock of 360 J was ineffective, we used biphasic shock of 360 J. Early recurrent atrial fibrillation (ERAF) was defined as a relapse of atrial fibrillation within 2 min after a successful cardioversion, acute recurrent - within 24 h. Two study groups (Bi vs Mo) did not differ with regard to age, body mass index, duration of AF episode (mean 98+/-147 days for the Bi group and 80+/-93 days for the Mo group, P=0.26), underlying heart disease, left atrial diameter, left ventricular ejection fraction. In the Mo group, more patients used amiodarone (59.82% vs 41.97%, P=0.002), in the Bi group more patients used propafenone (16.07% vs 8.93%, P=0.033). Cardioversion success rate was 97.32% in the Bi group and 79.46% in the Mo group (P<0.001). After biphasic shock of 360 J in Mo group, the cumulative success rate was 99.11%. Mean delivered energy and mean number of shocks were significantly lower in the Bi group (198.5+/-204.4 J, 1.5+/-0.9 shocks vs 489.1+/-464.2 J, 2.4+/-1.5 shocks). The efficacy of first shock was 66.96% in the Bi group and 37.5% in the Mo group (P<0.0001). Incidence of ERAF was 4.46% in both groups. Paddle position had no impact on efficacy of cardioversion and ERAF. For the cardioversion of atrial fibrillation, biphasic shock waveform has a higher success rate than monophasic shock waveform. We did not observe the influence of paddle positions on efficacy of cardioversion. Shock waveform and paddle position had no impact on ERAF. We recommend starting with biphasic energy of 150 J and monophasic of not less than 200 J for cardioversion of atrial fibrillation.