Abstract： Objective To explore the clinical value of dual-energy CT angiography (DE-CTA) in the diagnosis of carotid cavernous fistula. Methods Fourteen patients suspected of carotid cavernous fistula underwent DE-CTA between Dec. 2008 and Feb. 2010. Image post-processing of DE-CTA was performed with the dedicated software to obtain bone-removal and non bone-removal images. Four experienced radiologists evaluated image quality of DE-CTA, the number, location and size of fistulae and dilation of cerebral veins. The bone-removal and non bone-removal images were compared for displaying of fistulae and other related lesions. Kappa test was used to test the consistency of image quality evaluation between two radiologists. The sizes of fistulae were measured with bone-removal DE-CTA images and non bone-removal DE-CTA images respectively and their results were compared by using student's t-test and the correlation test. Results Twenty-eight intracranial internal carotid arteries of 14 patients showed acceptable image quality. Fourteen cases were diagnosed as carotid cavernous fistula with both bone-removal and non boneremoval DE-CTA images. Two patients presented with bilateral lesions, while 7 patients presented with left lesions and 5 patients had right lesions. Totally, sixteen fistulae lesions were detected in 14 patients. The mean size of fistulae was (0. 36 ±0. 10) mm by bone-removal images, whereas (0. 35 ±0. 11 ) mm by non bone-removal images. There was no significant difference between the two methods ( t = 0. 29, P ＞ 0. 05 ),and good correlation was found between the two methods ( r = 0. 97, P ＜0.05). Thirteen dilated cerebral veins in 8 cases were found by bone-removal images versus 9 dilated cerebral veins in 6 cases by non boneremoval images. Conclusions Image post-processing of DE-CTA is a simple and useful method to identify carotid cavernous fistula. DE-CTA is also useful for therapeutic planning.