Background: Despite extensive research to identify biomarkers of response in patients with non-muscle-invasive bladder cancer (NMIBC), there is no biomarker to date that can serve this purpose. Herein, we report how we leveraged serial urine samples to query a panel of cytokines at varying time points in an attempt to identify predictive biomarkers of response in NMIBC. Methods: Serial urine samples were collected from 50 patients with intermediate-or high-risk NMIBC enrolled in a phase II study, evaluating intravesical BCG AE intradermal HS-410 therapy. Samples were collected at baseline, week 7, week 13, week 28, and at end of treatment. A total of 105 cytokines were analyzed in each sample. To predict outcome of time-to-event (recurrence or progression), univariate and multivariable Cox analyses were performed. Results: Fifteen patients developed recurrence and 4 patients progressed during the follow-up period. Among clin-icopathologic variables, ever-smoker versus nonsmoker status was associated with an improved response rate (HR 0.38; 95% confidence interval (CI), 0.14-0.99; P ¼ 0.04). In the most clinically relevant model, the percent change (for 100 units) of IL18-binding protein-a (HR 1.995; 95% CI, 1.16-3.44; P ¼ 0.01), IL23 (HR 1.12; 95% CI, 1.01-1.23; P ¼ 0.03), IL8 (HR 0.27; 95% CI, 0.07-1.08; P ¼ 0.06), and IFNg-induced protein-10 (HR 0.95; 95% CI, 0.91-0.99; P ¼ 0.04) at week 13 from baseline best predicted time to event. Conclusions: Urinary cytokines provided additional value to clinicopathologic features to predict response to immunemodulating agents in patients with NMIBC. Impact: This study serves as a hypothesis-generating report for future studies to evaluate the role of urine cytokines as a predictive biomarker of response to immune treatments.