This study compared the predictive power of the systemic immune-inflammation index (SII) and Naples prognostic score (NPS) in determining the severity of coronary artery disease (CAD). The study included 1138 patients who underwent coronary computed tomographic angiography (CCTA). The primary outcome was the evaluation of CAD severity, determined by the Coronary Artery Disease-Reporting and Data System (CAD-RADS) obtained from the CCTA scans. A basic statistical model including age, gender, chest pain, diabetes mellitus, hypertension, hyperlipidemia, and smoking was built, and categorical variables, NPS (Naples 3,4 vs 0,1,2) and SII, were added to the basic statistical model. The net benefits of the predictive parameters were determined by a decision curve analysis, and the association between CAD-RADS and NPS, SII was quantified by odds ratios (OR) and 95% confidence intervals (CI). The decision curve analysis showed that adding SII to the statistical model had a better full range of probability of clinical net benefit compared with the baseline model (OR: 5.77, 95% CI 4.15–8.02, P < .001). However, adding the NPS ( P = .11) to the model did not outperform the basic statistical model. In conclusion, the SII may have a net predictive effect on top of traditional risk factors.