Objective: To investigate correlations between adenosine triphosphate chemotherapy response assay (ATP-CRA) and clinical outcomes after ATP-CRA-based chemotherapy for drug selection in patients receiving intravesical chemotherapy to prevent recurrence of superficial bladder cancer after surgery. Methods: The chemosensitivities of 12 anticancer drugs were evaluated, including 5-Fu ADM, and EPI, using ATP-CRA and primary tumor cell culture in 54 patients. In addition, a further 58 patients were treated according to clinical experience. Differences in post-chemotherapeutical effects between drug sensitivity assay and experience groups were compared. Results: The evaluable rate of the test was 96.3%, the clinical effective rate was 80.8%, the sensitivity rate was 97.6% (41/42), the specificity was 20%, the total predicting accuracy was 74.3%, the positive predictive value was 83.7% (41/49), the negative predictive value was 66.7% (2/3); in the drug sensitivity test group, the clinical effective rate was 80.8%, the experience group response rate was 63.8%, with a significant difference in clinical effects between the ATP-based sensitivity and experience groups (${chi}^2$=7.0153, P<0.01). Conclusion: ATP-CRA is a stable, accurate and potentially practical chemosensitivity test providing a predictor of chemotherapeutic response in patients with superficial bladder cancer.