Background The coronary artery disease reporting and data system (CAD-RADS) is designed for a uniform standardization of coronary computed tomography angiography (CCTA) reporting and further management recommendations of coronary artery disease (CAD). This study aimed to assess clinical validity, applicability, and reproducibility of CAD-RADS in the management of patients with CAD. Methods and results A single-center prospective study included 287 patients with clinically suspected or operated CAD who underwent CCTA. Four reviewers evaluated the CCTA images independently and assigned a CAD-RADS category to each patient. The invasive coronary angiography (ICA) was used as the reference standard for calculating diagnostic performance of CAD-RADS for categorizing the degree of coronary artery stenosis. The intra-class correlation (ICC) was used to test the inter-reviewer agreement (IRA). Reporting was provided to referring consultants according to the CAD-RADS. Based on ICA results, we have 156 patients with non-significant CAD and 131 patients with significant CAD. On a patient-based analysis, regarding those patients classified as CAD-RADS 4 and CAD-RADS 5 for predicting significant CAD, the CAD-RADS had a sensitivity, specificity, and an accuracy of 100%, 96.8 to 98.7%, and 98.3 to 99.3%, respectively, depending on the reviewer. There was an excellent IRA for CAD-RADS categories (ICC = 0.9862). The best cutoff value for predicting significant CAD was > CAD-RADS 3. Eighty-seven percentage of referring consultants considered CAD-RADS reporting system to be “quite helpful” or “completely helpful” for clinical decision-making in CAD. Conclusion CAD-RADS is valuable for improving CCTA structural reports and facilitating decision-making with high diagnostic accuracy and high reproducibility.