Multiple randomized clinical trials have demonstrated that atrial or dual-chamber pacing prevents paroxysmal and permanent atrial fibrillation (AF) in patients with symptomatic bradycardia as the primary indication for cardiac pacing. The benefit of atrial pacing for the prevention of AF is observed predominantly in patients with sinus node dysfunction. Emerging evidence also suggests that the risk of AF is directly linked to the proportion of time that ventricular pacing occurs. Consequently, pacemakers should be programmed to minimize the amount of ventricular pacing in patients with intrinsic atrioventricular conduction. Temporary atrial pacing following heart surgery may be of benefit for the prevention of perioperative AF. Atrial pacing has not been shown to prevent AF in patients without symptomatic bradycardia. In addition, selective pacing algorithms designed to prevent AF have minimal or no incremental benefits for the prevention of AF. At present, the role of selective atrial lead site(s) for the prevention of AF remains uncertain.