PURPOSE: To evaluate the influence of different reconstruction intervals of retrospectively electrocardiographically (ECG)-gated multi-detector row computed tomographic (CT) angiography on image quality of different segments of various types of coronary artery bypass grafts. MATERIALS AND METHODS: Twenty consecutive patients with 62 grafts underwent retrospectively ECG-gated four-channel multi-detector row CT angiography and conventional coronary angiography. Raw helical CT data were reconstructed at 0%-90% of the cardiac cycle in increments of 10%. Each graft was separated into three segments (proximal segment, graft body, and distal anastomosis). Three graft types were identified according to site of distal anastomosis. Two readers assessed image quality of segments and graft types. Effective radiation dose was calculated. RESULTS: Best image quality of all segments was obtained at a reconstruction interval of 50%-70% of the cardiac cycle. Image quality of the proximal segment did not vary significantly with different reconstruction intervals (analysis of variance, P =.8), whereas image quality of the graft body and distal anastomosis changed significantly with varying reconstruction intervals (P <.001). Distal anastomosis and body of types 1 and 2 grafts were best seen at 60%-70% of the cardiac cycle, whereas distal anastomosis and body of type 3 grafts were best visualized at 50%. Accuracy of CT angiography for detection of graft patency was 94% for reader 1 and 95% for reader 2. Effective dose for CT was 11.4 mSv for both men and women. Mean effective dose for angiography was 2.1 mSv for men and women. CONCLUSION: Optimal selection of reconstruction interval improves image quality of the graft body and of distal anastomosis in particular.