We wanted to compare the image quality and radiation dose of retrospective versus prospective ECG-gated dual-source CT imaging in pediatric patients with congenital heart diseases. We retrospectively evaluated 44 thorax CT images that were obtained using two different protocols (retrospective helical ECG-gated [n = 22] versus prospective non-helical ECG-triggering [n = 22]) in patients younger than 5 years. Two radiologists evaluated the image quality, using a 3- or 4-point scale, in terms of the degree of homogeneity of vascular enhancement, the stair-step artifacts, the subjective image noise and the overall subjective image quality. We compared the mean score of image quality, the frequency of marked stair-step artifact and the estimated radiation dose. The mean score of the homogeneity of vascular enhancement and the stair-step artifacts tended to be higher for the patients in the retrospective group than that for the patients in the prospective group, but this did not reach statistical significance. However, marked stair-step artifacts were significantly more frequent in the prospective scan group than that in the retrospective scan group (25.8 vs. 7.6%, respectively, P < 0.01). The mean scores for subjective image noise and the overall image quality were better in the retrospective group than those scores in the prospective group (P < 0.01). The mean estimated effective dose was higher for the retrospective ECG-gated helical scan than that for the prospective non-helical scan (0.79 vs. 0.21 mSv, respectively, P < 0.01). Retrospective ECG-gated helical dual-source CT using aggressive ECG pulsing can be obtained with an acceptable radiation dose and superior image quality, as compared to the prospective ECG-triggering technique, for the evaluation of congenital heart diseases in pediatric patients who are younger than 5 years old.