PURPOSE: To describe the MRI findings of cortical dysplasia (CD) according to pathological grading. MATERIALS AND METHODS: MRI findings in 24 patients with pathologically proven CD were retrospectively reviewed and weredescribed according to their histologic grading. The group consisted of 11 men and 13 women, and their median agewas 21.4(range, 5 to 41) years. Histologic findings were assigned one of three grades : Grade I (dyslamination ofcortical layers only); Grade II (additional dysplastic neurons); or Grade III(additional balloon cells). T1 and T2Wspin-echo MR images were obtained and were evaluated with specific reference to detection rate, location of theanomaly, the presence or absence of prolonged T2 relaxation in the underlying white matter, and configuration ofthe dysplastic cortex. RESULTS: Nine patients were found to be suffering from grade I CD, seven from grade II,and eight from grade III. MRI findings were abnormal in 58% of cases(14/24);three of these were grade I, four weregrade II, and seven were grade III. Among these 14 patients, four had the lesion in the frontal lobe;five, in thetemporal lobe;one, in the parietal lobe; two, in the sylvian fissure area, and two with tuberous sclerosis hadmultiple lesions. Of twelve patients with cortical dysplasia other than tuberous sclerosis, as seen on MRI, sixshowed poor gray-white matter differentiation, and six showed cortical thickening ; this ranged from 7-10mm inthickness with irregular, bumpy cortical surfaces. Areas of prolonged T2 relaxation in the underlying white matterwere present in six patients. CONCLUSION: MRI findings were abnormal in 58% of CD patients. The higher thepathological grading, the higher the detection rate of CD on MRI. To increase the detectability of slight corticaldysplasia, technical development, including new MR pulse sequences or imaging methods, is needed.