The complement system has been implicated in the pathogenesis of atherosclerosis. In addition, complement activation and complement-mediated brain injury have been found in a variety of central nervous system diseases, including stroke. However, there are limited data about the value of complement components for prediction of stroke. Complement C3 and C4 levels, in addition to a variety of established biomarkers, were measured in 389 men with known or suspected coronary artery disease referred for coronary angiography for a variety of indications at a Veterans Affairs Medical Center. Patients were followed prospectively for the development of stroke, which was defined and classified according to criteria of the Trial of Org 10172 in Acute Stroke Treatment (TOAST). All strokes were confirmed with magnetic resonance imaging or computed tomography. For the 97% of patients in whom 24-month follow-up data were available, there were 23 strokes (5.9%). By multivariate Cox proportional hazard analysis, complement C4 was an independent predictor of stroke, with a hazard ratio of 1.57 (95% confidence interval 1.03 to 2.39, p = 0.0358). The 24-month stroke-free survival rate for the patients whose complement C4 levels were equal to or below the median value for the entire cohort was 96.1% compared with 90.1% for patients whose complement C4 levels were above the median value, p = 0.0127 by log rank test). In conclusion, in a cohort of men across a spectrum of risk referred for coronary angiography, a single baseline determination of serum complement C4 level is an independent predictor of the future development of stroke.