We aimed to compare recurrence-free survival (RFS) and progression-free survival (PFS) of the patients with pure high-grade (HG) vs mixed-grade (MG) nonmuscle-invasive bladder cancer who received adequate bacillus Calmette-Guérin therapy. We conducted a retrospective cohort analysis using data from an institutional database. The study included patients diagnosed with HG nonmuscle-invasive bladder cancer at the initial transurethral resection specimen between 2010 and 2020. The initial transurethral resection specimens of all patients were reevaluated by a dedicated uropathologist. The percentage of low-grade tumor areas accompanying HG areas was determined for each case. Time-to-event analysis was performed using the Kaplan-Meier method. RFS and PFS rates were compared between groups. Of the 203 patients enrolled in the study, 69 (34%) had MG tumors. Recurrence was observed in 41 out of 134 patients (30.6%) in the HG group and in 19 out of 69 patients (27.5%) in the MG group. The 36-month RFS rates were 69% (CI: 62-77) and 72% (CI: 62-83) for the HG-urothelial carcinoma (UC) and MG-UC groups, respectively. The RFS rates were similar between groups (log-rank, P = .58). Progression was observed in 22 out of 134 patients (16.4%) in the HG group and in 4 out of 69 patients (5.8%) in the MG group. The 36-month PFS rates were 84% (CI: 77-90) and 94% (CI: 89-100) for the HG-UC and MG-UC groups, respectively. The pure HG-UC group had a worse PFS than the MG-UC group (log-rank, P = .042). Multivariate analysis demonstrated that age and tumor grade were significant risk factors for the development of progression. The indication of MG-UC category separately from pure HG carcinomas in the pathology report seems to be an important issue that can guide patient management. In this way, both more accurate risk classification and more accurate patient counseling can be performed. More importantly, the treatment plan can be made more accurately. For more precise conclusions, our results should be supported by prospective studies with larger sample size.