Partially due to recent reports that cardiac antiarrhythmic therapy may have adverse effects on patient survival, clinicians have become more interested in the nonpharmacological prevention of atrial fibrillation. There is a large body of literature that suggests that the rate of development of atrial fibrillation in paced sick sinus syndrome patients is much lower in those patients who have received an atrial-based system, rather than a VVI system. However, all the published studies to date are retrospective, and fraught with potential bias favoring the AAI or DDD group. The authors strongly believe that the only way to determine if these suggestive but uncertain retrospective analyses are correct is to apply the same scientific rigor to this problem as has been applied to many other problems in cardiovascular medicine and perform a prospective randomized trial. A proposed trial design is discussed.