Discovery of genes that influence stroke risk might be facilitated if phenotypic heterogeneity of the population studied were reduced by studying individual subtypes of stroke. Standardized systems for classifying stroke are generally mechanistic or syndromic. Mechanistic systems (eg, Trial of ORG10172 in Acute Stroke Treatment) are based on presumed pathophysiologic basis of occlusion. Syndromic systems (eg, Oxfordshire Community Stroke Project) are based on signs and symptoms of neurologic dysfunction. Linkage and candidate gene studies should use such standardized systems in the search for subtype-specific genetic risk factors to increase the validity of interstudy comparisons.