BACKGROUND: Two classification systems exist for subtypes of acute cerebral infarction. One was developed for the Trial of Org 10172 in Acute Stroke Treatment (TOAST), based primarily on etiology. The other is the Oxfordshire Community Stroke Project (OCSP), based on clinical features. OBJECTIVE: To evaluate the relationship between OCSP and TOAST classifications in terms of stroke location and etiology in 126 patients with acute ischemic stroke confirmed by transcranial magnetic resonance imaging (MRI). DESIGN, TIME AND SETTING: Retrospective case analysis. Transcranial MRI, diffusion weighted imaging, and magnetic resonance angiography were performed in 126 patients with acute stroke during the first 48 hours following admission to the Department of Neurology, Navy General Hospital of Chinese PLA. PARTICIPANTS: A total of 126 patients with acute stroke, comprised of 71 males and 55 females, admitted to the Navy General Hospital of Chinese PLA between December 2005 and April 2006 were included. METHODS: Of 126 patients with acute stroke, 13 exhibited total anterior circulation infarcts (TACI), 51 had partial anterior circulation infarcts (PACI), 28 suffered posterior circulation infarcts (POCI), and 34 had lacunar infarcts (LACI) based on OCSP classification. However, according to TOAST classification, 19 cases were a result of large-artery atherosclerosis, 32 by cardioembolism, 36 by small-vessel occlusion, 1 by stroke of other determined etiology, and 38 by stroke of undetermined etiology. MAIN OUTCOME MEASURES: The corresponding relationship of the subtypes of acute stroke based on OCSP and TOAST. RESULTS: Of patients with TACI, 8 (61.5%) were caused by cardioembolism. Of patients with PACI, 16 (31.4%) were caused by large-artery atherosclerosis and 17 (33.3%) by cardioembolism. Of patients with POCI, 12 (42.8%) were a result of small-vessel occlusion. Of patients with LACI, 17 (50.0%) were caused by hypertension and arteriolar sclerosis. CONCLUSION: (1) The OCSP system is related to anatomical and pathophysiological processes, and can correctly identify the size, location, and cause of cerebral infarcts. It can also act as a reference for prognosis estimation and recurrence prevention of ischemic stroke. (2) The stroke of undetermined etiology occupies the largest percent in the subtypes of stroke, according to TOAST. (3) The classification of OCSP exhibits a close relationship to TOAST. One was developed for the Trial of Org 10172 in Acute Stroke Treatment (TOAST), based primarily on etiology. The other is the Oxfordshire Community Stroke Project (OCSP), based on clinical features. OBJECTIVE: To evaluate the relationship between OCSP and TOAST classifications in terms of stroke location and etiology in 126 patients with acute ischemic stroke confirmed by transcranial magnetic resonance imaging (MRI). DESIGN, TIME AND SETTING: Retrospective case analysis. Transcranial MRI, diffusion weighted imaging, and magnetic resonance angiography were performed in 126 patients with acute stroke during the first 48 hours following following the Department of Neurology, Navy General Hospital of Chinese PLA. PARTICIPANTS: A total of 126 patients with acute stroke, comprised of 71 males and 55 females, admitted to the Navy General Hospital of Chinese PLA between December 2005 and April 2006 were METHODS: Of 126 patients with acute stroke, 13 had total anterior back infarcts (TACI), 51 had partial anterior circlip infarcts (PACI), 28 suffered posterior frontal infarcts (POCI), and 34 had lacunar infarcts (LACI) based on OCSP classification. However, according to TOAST classification, 19 cases were a result of large-artery atherosclerosis, 32 by cardioembolism, 36 by small-vessel occlusion, 1 by stroke of other determined etiology, and 38 by stroke of undetermined etiology. MAIN OUTCOME Of patients with PACI, 16 (31.4%) were caused by large-artery Of patients with POCI, 12 (42.8%) were a result of small-vessel occlusion. Of patients with LACI, 17 (50.0%) were caused by hypertension and arteriolar sclerosis. CONCLUSION: ( 1) The OCSP can be used as a reference for prognosis estimation and recurrence prevention of ischemic stroke. (2) The stroke of undetermined etiology occupies the largest percent in the subtypes of stroke, according to TOAST. (3) The classification of OCSP exhibits a close relationship to TOAST.