Recommendations for identifying patients with nonvalvular atrial fibrillation who are at low, moderate, or high risk for thromboembolism have recently converged, advocating almost identical risk-stratification criteria (1–3). The widely promulgated CHADS2 scheme uses a system that assigns 1 point each for congestive heart failure, hypertension, age 75 years or older, and diabetes mellitus and 2 points for previous stroke or transient ischemic attack (4). Anticoagulation is generally not recommended for patients with atrial fibrillation who are at low risk for thromboembolism (CHADS2 score of 0; about 20% of patients with atrial fibrillation, who face an average stroke risk of 1% per year), but it is favored for patients at high risk (CHADS2 score ≥2; about 45% of patients with atrial fibrillation, whose stroke risk averages 4% to 5% per year) who can tolerate warfarin (1–3). For those at moderate risk (CHADS2 score of 1; about 35% of patients with atrial fibrillation, with an average stroke risk of 2% per year), either warfarin or aspirin is recommended (1), with warfarin being preferred in a prominent guideline (3). In this issue, Singer and colleagues (5) question the net clinical benefit of adjusted-dose warfarin for at least half of patients with atrial fibrillation, including the one third deemed to have moderate stroke risk (CHADS2 score of 1), who frequently receive warfarin now.