Atrial fibrillation (AF) is not a benign condition. Although AF is the most common sustained cardiac arrhythmia, the management of AF remains controversial. In recent years, several clinical trials comparing rhythm control with rate control for AF management found that rhythm control offers no advantage over rate control, but the findings may not apply equally to all patients. The main point is the high risk of AF recurrence in these trials and the toxic effects and poor efficacy of presently available antiarrhythmic drugs negate the benefit of rhythm control. Rhythm control should be considered as the preferred initial therapy if sinus rhythm had been maintained in a higher proportion of patients. For the moment, the optimal strategy to treat AF should be individualized, with a bias towards rate control if patients are at high risk for arrhythmias or can be kept asymptomatic. Rhythm control may still be justified in younger patients or patients who still have symptoms despite adequate rate control. In addition, patients with new or first-episode AF often warrant at least an initial trial of rhythm control.