Cardiac CT angiography (cCTA) has become an established method for the assessment of congenital heart disease. However, the potential harmful effects of ionizing radiation must be considered, particularly in younger, more radiosensitive patients. In this study, we sought to assess the temporal change in radiation doses from pediatric cCTA during an 8-year period at a tertiary medical center. This retrospective study included all patients ≤18 years old who were referred to electrocardiography (ECG)-gated cCTA for the assessment of congenital heart disease or inflammatory disease (Kawasaki disease) from November 2004 to September 2012. During the study period, 95 patients were scanned using 3 different scanner models—64-slice multidetector CT (64-MDCT) and first- (64-DSCT) and second-generation (128-DSCT) dual-source CT—and 3 scan protocols—retrospective ECG-gated helical scanning (RG), prospective ECG-triggered axial scanning (PT), or prospective ECG-triggered high-pitch helical scanning (HPH). Effective dose (ED) was calculated with the dose length product method with a conversion factor (k) adjusted for age. ED was then compared among scan protocols. Image quality was extracted from clinical cCTA reports when available. Overall, 94 % of scans were diagnostic (80 % for 64-slice MDCT, 93 % for 64-slice DSCT, and 97 % for 128-slice DSCT).With 128-DSCT, median ED (1.0 [range 0.6–2.0] mSv) decreased by 85.8 % and 66.8 % compared with 64-MDCT (6.8 [range 2.9–13.6] mSv) and 64-DSCT (2.9 [range 0.9–4.1] mSv), respectively. With HPH, median ED (0.9 [range 0.6–1.8] mSv) decreased by 59.4 % and 85.4 % compared with PT (2.2 [range 0.9–3.4] mSv) and RG (6.1 [range 2.5–10.6] mSv). cCTA can now be obtained at very low radiation doses in pediatric patients using the latest dual-source CT technology in combination with prospective ECG-triggered HPH acquisition.