Randomized clinical trials have proved that warfarin therapy decreases the risk of stroke in patients with nonvalvular atrial fibrillation and in those who have had a myocardial infarction. In patients who are not candidates for long-term anticoagulant therapy, aspirin is beneficial, but the reduction in risk is smaller with aspirin than with warfarin. In patients with cerebral ischemic symptoms of noncardiac origin, aspirin and ticlopidine reduce the risk of stroke, but the benefit is modest. Given alone, neither dipyridamole nor sulfinpyrazone prevents stroke. The question remains whether either of these drugs plus aspirin is better than aspirin alone. The optimal dose of aspirin for stroke prevention has not been established. Carotid endarterectomy reduces the risk of stroke in symptomatic patients with at least 70 percent stenosis, as determined by arteriography. Current trials are addressing the question of whether endarterectomy is beneficial for patients with moderate degrees of carotid stenosis. The benefit of endarterectomy for patients with asymptomatic carotid lesions remains unclear.