To evaluate the efficacy of flecainide acetate in the prevention of paroxysmal atrial fibrillation and flutter, 43 patients (23 men) (mean age 53 years) were randomized blindly to receive either placebo or 150 mg of flecainide twice per day for consecutive periods of 3 months. Attacks were verified by a minielectrocardiogram event recorder. If intolerable symptoms developed, the protocol allowed patients to cross over between treatments before the end of the first 3-month period. Four patients crossed over prematurely, between 1 week and 1 month, and 15 between 1 month and 3 months. The remaining 24 patients completed both 3-month periods. In all 3 treatment intervals, there was a significant reduction in the number of attacks during flecainide treatment (p less than 0.002). Complete suppression was seen in 15 of 43 patients (35%) treated with flecainide for 1 week, in 18 of 39 (46%) treated for 1 month and in 12 of 24 (50%) completing all 3 months in each period. Adverse effects were reported in 32 of the 43 patients (74%) treated with flecainide, but only 2 were withdrawals. One patient died suddenly. In comparison, 3 of 43 patients (7%) reported adverse effects in the placebo group. In conclusion, flecainide significantly suppressed the number of attacks of paroxysmal atrial fibrillation and flutter. Adverse effects were frequent but were mostly tolerable.