In symptomatic patients undergoing coronary CTA for suspected coronary artery disease (CAD), we assessed if quantification of plaque burden, in addition to luminal narrowing and clinical risk factors, offers incremental value for the identification of ischemic CAD on a per patient level. We evaluated 2145 patients who underwent coronary CTA for suspected CAD with sequential selective downstream ¹⁵O-water positron emission tomography (PET) myocardial perfusion imaging. Coronary CTA scans were analyzed using Artificial Intelligence-guided Quantitative Computed Tomography (AI-QCT), with measurement of maximum diameter stenosis, percent atheroma volume (PAV), percent calcified plaque volume (CPV) and percent noncalcified plaque volume (NCPV). Ischemic CAD was defined as the presence of abnormal stress perfusion on ¹⁵O-water PET. PAV on top of the clinical variables and ≥ 50% stenosis improved the prediction of ischemic CAD on a per patient level as compared to clinical variables and ≥ 50% stenosis (AUC = 0.91 vs. AUC = 0.87, p < 0.001). The best diagnostic performance was achieved when PAV with a cut-off value of 12.2% was applied in patients with intermediate (30-70%) stenosis; using this approach, the sensitivity, specificity, positive and negative predictive values and diagnostic accuracy for ischemic CAD were 76%, 91%, 64%, 95% and 88%. The addition of quantitative plaque volume on top of clinical variables and ≥ 50% diameter stenosis improves the detection of ischemic CAD as defined by PET perfusion imaging. Applying a PAV threshold of 12.2% in patients with intermediate stenosis provided the best diagnostic performance.