The aim of the study was to evaluate the influence of magnesium sulfate on the efficacy and safety of pharmacological cardioversion with nibentan (NB) in doses up to 0,125 mg/kg in patients with persistent atrial fibrillation (AF) and flutter (AFL). Calculated dose of NB was 0.125 mg/kg. It was administered as 2 bolus injections (0.0625 mg/kg each) performed with the interval of 15 minutes. The study included 64 patients (pts) (45 male, age 54+/-9,9 years) with persistent AF (n=56) and AFL (n=8) with median arrhythmia duration 6,7+/-6,8 months (8 days to 36 months). Pts were divided into two groups. In the first (I) group NB was used without preliminary magnesium sulfate administration, in the second group (II) magnesium sulfate was injected in a dose of 50 mg/kg, followed by 0.83 mg/kg/min infusion before NB administration. There was no difference between groups in conversion rates of atrial arrhythmias: 74% and 69%, in groups I and II, respectively (p>0.4). Mean effective dose of NB was the same in investigated groups - 0.09 mg/kg. Administration of initial dose of NB (0.0625 mg/kg) converted AF/AFL to sinus rhythm within 15-min interval in 29.7% of patients. Efficacy of full dose of NB (0.125 mg/kg) was 100% in AFL, 68% in AF. No side effects were registered after initial injection of 0.0625 mg/kg. After administration of 2 boluses (0.125 mg/kg) "torsade de pointes" developed in 2 pts (3%). Magnesium sulfate administration had no influence on NB efficacy and rate of its proarrhythmic events. Limitation of NB total dose to <or= 0.125 mg/kg divided into 2 bolus injections does not lead to loss of antiarrhythmic efficacy and makes its administration more safe.