In recent-onset atrial fibrillation, intravenous propafenone has been shown to effectively restore sinus rhythm, whereas the efficacy of intravenous digoxin has been questioned. We directly compared these 2 drugs and placebo in acute atrial fibrillation. One hundred twenty-three patients with atrial fibrillation lasting <72 hours were randomized to a 10-minute intravenous infusion of either propafenone (2 mg/kg, 41 patients) or digoxin (0.007 mg/kg, 40 patients) or placebo (42 patients). After 1 hour, nonconverted propafenone or digoxin patients were switched to the alternative drug, while nonconverted placebo patients were randomized to either propafenone or digoxin. The observation time ended 1 hour later. By 1 hour, conversion rates were 49% in the propafenone group, 32% in the digoxin group (p = 0.12), and 14% in placebo group (p <0.001 vs propafenone, p = 0.08 vs digoxin). After crossover, digoxin converted 5% of propafenone patients, while propafenone converted 48% of digoxin patients (p <0.05). In the 36 nonconverted placebo patients, sinus rhythm was obtained in 53% of cases with propafenone, and in 5% with digoxin (p <0.05). Globally, among the 116 patients who received a drug as first treatment, 30 of 60 patients (50%) were converted by propafenone versus 14 of 56 (25%) by digoxin (p <0.01) (odds ratio 2.0, 95% confidence interval 1.19 to 3.36). In nonconverters, the ventricular rate reduction was faster (15 vs 45 minutes) and more prominent (−24% vs −14%) with propafenone than with digoxin. In conclusion, intravenous propafenone terminates atrial fibrillation more effectively than either placebo or intravenous digoxin. In addition, in nonconverted patients, it obtains a more rapid and marked control of the ventricular rate.