This randomized, double-blind trial compared cardioversion rates between AZD7009 infusion (15-minute 3.25 mg/min, 15-minute 4.4 mg/min, or 30-minute 3.25 mg/min) and placebo infusion (15 or 30 minutes) in patients with atrial fibrillation (AF) scheduled for DC cardioversion. One hundred sixty-eight patients were randomized, 167 received study treatment, and 159 were included in perprotocol analyses. The mean duration of current AF episode was 47 days (range, 0.8-92). In the AZD7009 30-minute 3.25 mg/min group, 21 of 42 patients converted within 90 minutes, compared with 7 of 39, 7 of 36, and 0 of 42 patients in the 15-minute 3.25 mg/min, 15-minute 4.4 mg/min, and combined placebo groups, respectively. Patients not converted within 90 minutes underwent DC cardioversion. In patients with AF episodes </=30 days, conversion rates with AZD7009 30-minute 3.25 mg/min and after placebo followed by DC cardioversion were 82% and 83%, respectively. AZD7009 was generally well tolerated. Mean QTc increased by 15% to 20% in the AZD7009 treatment groups. One patient experienced a ventricular tachyarrhythmia of 7 beats, with features of torsades de pointes. AZD7009 was associated with a 50% conversion rate in the best-dose group. In patients with AF episodes <or=30 days, the conversion rates after AZD1305 and placebo followed by DC cardioversion were observed to be similar.