Background: Trying to reduce radiation exposure from cardiac imaging is mandatory while maintaining diagnostic image quality (IQ). Using a high-pitch spiral dual-source computed tomography (DSCT) protocol for left atrial (LA) mapping, we sought to determine IQ and radiation dose in patients with atrial fibrillation scheduled for radiofrequency ablation. Methods: Fifty-nine patients (29 women; mean age = 53 y) underwent CT angiography between 2013 and 2016: 26 patients with retrospective ECG-gated (classic) and 33 with high-pitch (Flash) protocols on a second-generation 128-DSCT system (SOMATOM Definition Flash). CT images of the LA were integrated into an electroanatomic system (NavX). Two independent blinded readers evaluated IQ using a 3-point scale and the LA contrast density. Dose-length product (DLP) was obtained from each patient protocol, and effective radiation dose (ERD) was calculated according to the European guideline for CT. Results: The rate of diagnostic IQ (score 3 or 2) was 87.9% for the flash group and 96.2% for the classic group, which was not significantly different between the 2 groups (P = 0.250). The results of objective IQ measurements showed that a central LA contrast density above 350 Hounsfield units (diagnostic) was present in 21 (80.8%) images in the classic group and 26 (78.8%) images in the flash group, which was not statistically different between the 2 groups (P = 0.850). There were significant differences (P < 0.001) in DLP and ERD between the 2 groups: the values were lower in high-pitch scan than in retrospective ECG-gated scan (151.30 ± 39.44 vs. 776.61 ± 243.63 and 2.11 ± 0.55 vs. 10.872 ± 3.41, respectively). Conclusions: High-pitch (Flash) DSCT is an acceptable CT angiography method for reducing radiation dose without compromising IQ for LA and pulmonary venous imaging in patients with atrial fibrillation.