OBJECTIVE To determine the clinical syndromes, etiopathogenesis, and prognostic factors in a prospectively evaluated multiethnic young stroke population. METHODS Only first-ever patients with a World Heath Organization definition of stroke and anatomic brain imaging were included. A hierarchy of investigative modalities divided into three tiers was applied and a range of standardized scales scored in each patient. This allowed quantification of clinical deficit, etiopathogenesis, disability, and handicap. Standardized stroke scales included the Canadian Neurological Scale (CNS), the Oxfordshire Community Stroke Project (OCSP) clinical stroke scale, and TOAST (Trial of Org 10172 in Acute Stroke Study) etiological classification. Disability was measured with the Barthel Index and handicap with the Rankin Scale; cognitive impairment was separately evaluated according to predefined criteria. A prognostication measure was made in some patients with the Cerebral Perfusion Index (CPI). RESULTS A total of 236 patients was evaluated of whom 64 were excluded because of no lesion consistent with stroke on brain scanning leaving 172 for analysis. There were 87 women, 85 men, with a mean age of 43.8 years (range, 15 to 49 years). Despite many different predefined symptoms, 38 patients (22%) could not be classified. Hypertension (31%) and smoking (19%) were the most commonly encountered risk factors, with more recently determined risk factors such as infection (6%) and emotional stress (5%) relatively frequent. With respect to etiology, the TOAST category "other" was the most numerous group, numbering 93 of 172 (55%) with prothrombotic states in 25 (15%), vasculitis in 21 (12%), and dissection in 12 (7%) being the most frequent causes. Proportions of the remaining categories were small vessel disease (16%), cardioembolism (13%), large vessel disease (10%), and unknown (6%). X-square analysis for an association between the clinical OCSP and TOAST classifications was not significant. Severity of stroke was generally mild as judged by the CNS and Rankin scales. A high proportion of patients had cognitive impairment (54%). A cerebral perfusion index was possible in 31 patients, most of whom had a medium prognosis. CONCLUSIONS in this hospital-based consecutive series, most young stroke patients in our region were grouped into nonatherogenic (mostly prothrombotic states, infection asssociated and dissection) and noncardiac causes with a definite or probable cause found in 94%. The wide variety of stroke symptoms recorded in this study underscores the heterogeneity of stroke presentation and caution in the emergent evaluation of patients. Cognitive impairment in the majority of stroke patients in the acute and subacute stroke period has important implications for degree of clinical deficit especially as it applies to stroke scales and treatment trials.