PURPOSE: The purpose of this study was to determine the role of CT and MR imagings in the diagnosis aortic dissection and differentiation between the true and false lumen. MATERIALS AND METHODS: We retrospectively studied forty patients with aortic dissection(AD) diagnosed imagings or surgery. Of the forty patients, 19 were examined with only CT, 14 with CT and MR, and 7 with MI~: Our points of view were(1) the classification of AD according to configuration of intimal flap by cross-sectional imaging, (2) differentiation between the true and false lumens, (3) the course of the false lumen, and (4)! detectability of the origin of major branch vessels of the abdominal aorta. RESULTS: The classification by corss-sectional imaging were crescentic(65%), circumferential(15%), flat(12%), and irregular(8%) type, in which false negative diagnosis was made in 1 case of crescentic and circumferential type, respectively. In 2 case of flat type and 1 case of irregular type, the differentiation between the true and false lumen was impossible with CT. The course of the false lumen in descending thoracic aorta revealed countrclock wise rotation(66%), clockwise rotation(5%) or fixed(29%) apperance. MR imaging was superior to CT in the detection of the origin of major branch vessels of the abdominal aorta. The determination of the origin of major branches of abdominal aorta arising from the true and false lumen were impossible in 2 cases in which only CT was done. CONCLUSION: Diagnosis of crescentic and circumferential types of AD with narrow and thrombosed false lumen was problematic in both CT and MR with no difference of diagnostic accuracy between the two modalities. The differentiation between the true and false lumen was difficult in flat and irregular types with only CT. Therefore, when surgical treatment is considered as in type B aortic dissection, MR imaging is recommended in order to determine the origin of major branch vessels.