We evaluated the association between alcohol consumption and carotid atherosclerosis in 261 consecutive patients greater than 50 years old admitted to our community-based primary-care center with first ischemic stroke; their characteristics were entered into a computerized data bank (Lausanne Stroke Registry). Reported regular alcohol consumption was compared with the presence and severity of internal carotid artery disease as assessed by duplex scanning with spectral analysis of the Doppler signal and real-time B-mode imaging at the level of the carotid bifurcation. We found an inverse linear relation between light-to-moderate alcohol intake (less than or equal to 4 standard drinks/day) and severity of internal carotid artery stenosis. No conclusion could be drawn for heavier drinkers because there were too few. A logistic regression model showed that hypertension, cigarette smoking, and age in men and diabetes mellitus and cigarette smoking in women strongly counterbalanced the potential benefit of alcohol consumption. Although regular alcohol drinking cannot be advocated on the basis of our findings, light-to-moderate consumption of alcohol is the first factor to be inversely associated with extracranial carotid atherosclerosis in symptomatic patients with cerebrovascular disease.