To the Editor: Pluymaekers et al. (April 18 issue)(1) evaluated the occurrence of spontaneous cardioversion in patients with recent-onset (<36 hours) atrial fibrillation. The proposed wait-and-see approach included cardioversion within 48 hours in the absence of spontaneous resumption of sinus rhythm, and delayed cardioversion was actually necessary in 28% of the patients randomly assigned to this strategy. We think that the idea of waiting for spontaneous cardioversion of atrial fibrillation is interesting, but for practical reasons, better patient targeting would be required. In two previous prospective, randomized trials, we found that in patients with recent-onset atrial fibrillation, spontaneous cardioversion can . . .