Background <p>Risk stratification is crucial for outcome comparison and standardization of treatment and follow-up in non-muscle-invasive bladder cancer (NMIBC).</p> Purpose <p>To investigate the feasibility of using CT imaging features to evaluate the risk stratification of NMIBC.</p> Materials and methods <p>Data from 168 patients pathologically diagnosed with NMIBC were retrospectively collected. Cases were stratified into low (<i>n</i> = 50), medium (<i>n</i> = 23), and high-risk groups (<i>n</i> = 95) according to the European Association of Urology guidelines. Preoperative CT imaging features were evaluated independently by two radiologists. Interobserver agreement was assessed using kappa and intraclass correlation coefficients. Univariate analysis was performed using Fisher’s exact test, the chi-square test, and ordinal logistic regression. Variables with significant associations were included in the generalized linear model (GLM).</p> Results <p>The GLM showed that tumor location, number, and long diameter helped evaluate risk stratification. Tumors located in the posterior wall (<i>p</i> = 0.004), side wall (<i>p</i> = 0.001), and ureteral orifice (<i>p</i> = 0.032) had a higher probability of being in the high-risk group. Tumors with longer maximal diameters tended to exhibit a higher risk (<i>p</i> = 0.005). The presence of multiple tumors had a probability of being in a relatively higher risk group than single tumors (<i>p</i> = 0.002). The model’s AUC values for predicting low, medium, and high-risk tumors are 0.83 [95% confidence interval (CI) 0.77–0.89], 0.67 (95% CI 0.56–0.79), and 0.70 (95% CI 0.62–0.78), respectively. The micro-average AUC (Micro-AUC) for predicting overall tumor risk stratification is 0.73 (95% CI 0.68–0.77).</p> Conclusion <p>It is feasible to evaluate the risk stratification of NMIBC using preoperative CT imaging features.</p>