The efficacy and safety of intravenous flecainide to convert recent-onset atrial fibrillation (AF) (present for greater than or equal to 30 minutes and less than or equal to 72 hours and a ventricular response greater than or equal to 120 beats/min) was investigated. A total of 102 patients without severe heart or circulatory failure were randomized to receive either intravenous flecainide (2 mg/kg, maximum dose 150 mg; 51 patients) or placebo (51 patients) in a double-blind trial. Digoxin (500 micrograms intravenously) was administered to all patients who had not previously been receiving digoxin. The electrocardiogram was monitored continuously during the study. In 29 (57%) patients stable sinus rhythm was restored within 1 hour after flecainide and in only 7 (14%) given placebo (chi square 18.9; p = 0.000013; odds ratio 8.3; 95% confidence interval 2.9-24.8). Reversion to sinus rhythm within 1 hour after starting the trial medication was considered a pretrial end point and likely to be due to a drug effect. At the end of the 6-hour monitoring period, 34 patients (67%) in the flecainide group were in sinus rhythm whereas only 18 (35%) in the placebo group had reverted (chi square 8.83, p = 0.003; odds ratio 3.67; 95% confidence interval 1.5-9.1). Significant hypotension, although short lived, was more common in the flecainide group. One patient given flecainide developed torsades de pointes and was successfully electrically cardioverted. Flecainide is useful for the management of recent-onset AF both for control of the ventricular response and conversion to sinus rhythm.(ABSTRACT TRUNCATED AT 250 WORDS)