Atrial fibrillation (AF) is the most common sustained arrhythmia, exacting a substantial toll in cardiovascular morbidity and mortality. Until recently, the prevailing philosophy has been that restoration and maintenance of normal sinus rhythm, as opposed to control of ventricular response rate, was the optimal approach to treatment of AF. A series of landmark trials (AFFIRM, RACE, STAF, and PIAF) have called this strategy into question, suggesting outcomes are equivalent with both approaches. These data do not mean that rhythm control is not beneficial, but highlight the limitations of current therapies to achieve and maintain sinus rhythm. Limitations of the rhythm-control strategy may be related to our difficulty in accurately documenting symptomatic benefit from this approach, the lack of efficacy and excessive adverse-effect burden associated with currently available antiarrhythmic agents, and selection biases in the enrollment of patients in clinical trials of rhythm control versus rate control, making the trials incompletely representative of the population eligible for therapy. New pharmacologic agents under development feature increased atrial selectivity or multichannel-blocking properties (or both). As a result, these compounds may be more effective in prolonging atrial refractoriness and may also have reduced proarrhythmic potential. It is premature to abandon the concept of rhythm control in AF until we have trials designed to include younger and highly symptomatic patients, more sensitive tools to measure symptomatic improvement, and safer, more effective antiarrhythmic agents.