The aim of this study was to compare the efficacy and safety of propafenone (2 mg/kg intravenously over 10 minutes) and digoxin (0.007 mg/kg intravenously over 10 minutes) in the conversion of recent onset (<72 hours) atrial fibrillation to sinus rhythm. A consecutive series of 50 uncomplicated patients (no contraindications to either drug and without clinical heart failure) was randomly divided into 2 groups of 25 patients each. The 2 groups were similar in age, gender, concomitant diseases, previous paroxysmal atrial fibrillation, duration of the arrhythmia, electrocardiographic parameters, basal blood pressure and 2D-echocardiographic profile. Sinus rhythm was restored in a higher percentage of propafenone-treated patients [22 of 25 patients (88%) vs 8 of 25 (32%); p = 0.00005] and within a shorter time period (mean 22 vs 96 minutes; p = 0.0004) relative to digoxin-treated patients. The study design allowed for conditional crossover of patients who did not respond to initial drug treatment. 11 of 17 patients who did not respond to digoxin were treated with propafenone; conversion to sinus rhythm was achieved in 6 of 11 (54%) patients within 3 hours. No clinically relevant adverse effects were noted in either treatment group. Neither treatment gave rise to any significant modification in QRS- or QTc-interval morphology. In the propafenone group, normal left atrium measurement values (anteroposterior diameter, longitudinal diameter, transversal diameter and a preserved ejection fraction) were considered predictive factors for conversion to sinus rhythm. Additionally the high percentage of early conversions obtained with propafenone allowed earlier discharges in this group of patients compared with that of digoxin. In conclusion, propafenone was well tolerated and was more effective than digoxin in the treatment of patients with recent onset atrial fibrillation. The drug offered advantages over digoxin with respect to its rapid onset of action and reduction in the number of patients who required hospitalisation.