Background: Electrical direct-current cardioversion (DCCV) has become a routine therapy for atrial fibrillation (AF), although some uncertainty remains regarding the optimal energy settings. Aims: This study examines whether the use of a higher initial energy monophasic shock of 360 joules (J) for external DCCV, in patients with persistent AF would prove more effective, yet as safe, as the use of a lower initial energy 200 J shock. Methods: A cohort of 107 patients with persistent AF was prospectively randomized to an initial synchronized DCCV shock of 360 J versus 200 J (n = 50 vs 57), followed by a similar shock sequence thereafter of four further shocks of 360 J for the two groups. In all patients the levels of troponin I (cTnI) were measured precardioversion and 18–20 hours later, the following day. In a subgroup of 36 patients in each group, the levels of creatine kinase (CK) and aspartate transaminase (AST) were measured pre- and 18–20 hours postcardioversion. Results: The success rate for DCCV was significantly higher in the 360 J group compared to the 200 J group (96.0% vs 75.4%, P = 0.003). The mean CK IU/L levels (1137.5.0 vs 2411.8, P = 0.014) and AST levels (39.83 vs 52.86, P = 0.010) were significantly lower in the 360 J group compared to the 200 J group. There was no statistical rise in cTnI (μg/L) in either group (P = 1.00). The average number of shocks delivered (1.84 vs 2.56, P = 0.006) was significantly less in the 360 J group than in the 200 J group, although total energy requirements for DCCV were similar for the two groups (662.4 J vs 762.4 J, P = 0.67). Conclusion: For patients with persistent AF the use of a higher initial-energy monophasic shock of 360 J achieves a significantly greater success rate, with less skeletal muscle damage (and no cardiac muscle damage) as compared with the traditional starting energy of a 200 J DC shock.