A population of 105 patients with recent onset (< 72 h) atrial fibrillation was randomly treated with propafenone as a single oral loading dose of 450 mg (Regimen A) or 600 mg (Regimen B) or with placebo. A 24-h Holter was performed. Criteria of efficacy were conversion to sinus rhythm at 2, 4 and 8 h compared to placebo and also significant reduction of mean ventricular rate in persistent atrial fibrillation. After 2 h, regimen B was more effective than either regimen A (43% vs. 8%; p = 0.001) or placebo (11%; p = 0.004). At 4 h, both the active treatments were more effective than placebo (17% vs. 46% regimen A and 57% vs. regimen B; p < 0.04 and p < 0.001, respectively). Sinus rhythm resumed within 24 h in 71%, 80% and 69% of the patients with regimen A, B and placebo, respectively (p = not significant). The mean ventricular rate reduction after 1 h was 8%, 11% and 4% for regimen A, B and placebo, respectively (p < 0.005 vs. regimen B), and 17%, 25% and 6% respectively (p < 0.001 placebo vs. regimen A and B, p < 0.05 regimen B vs. A) at 2 h. No major adverse effect occurred. Atrial flutter with 1:1 atrioventricular conduction only in one case who received placebo. Propafenone acute oral administration is more effective than placebo in rapidly converting recent-onset atrial fibrillation to sinus rhythm and may be the treatment of choice in this setting limiting hospitalization and contributing to improved quality of life.