Objective: To evaluate the feasibility and diagnostic accuracy of 64-slice spiral computed tomography(MSCT) in differentiating the stability of plaque in coronary arteries. Methods: Seventy-six patients diagnosed of having CAD clinically had undergone 64-MSCT angiography. They were divided into two groups according to clinical materials: ACS and SA. The number, morphology and CT value of plaques obtained by MSCT were compared between the two groups, as well as the change of CT value after delayed scanning. The remodeling index had also been analyzed. Results: The number of plaques did not differ between the two groups, however, in the culprit lesion of ACS group, the minimum CT density[(29±24)HU] was lower than that of SA group[(69±29)HU](P<0.0001). Similarly, in ACS patients with multiple plaques, the minimum plaque density[(22±25)HU] was lower in the culprit coronary segment than in the non-culprit segment[(50±28)HU](P<0.002). The reduction of CT density in SA group was significantly lower than ACS group after delayed scanning. With respect to remodeling index, the rate of positive remodeling in ACS group was 58.1%, while this rate was 24.7% in SA group. The chi-test demonstrated significant difference between the two groups. Conclusion: 64-MSCT can sufficiently detect vulnerable plaques from coronary plaques in patients with coronary artery disease through determining their density and morphology et al, but its clinical use may presently be limited due to image quality and evaluation means to a certain extent.