Papillary urothelial neoplasms with deceptively bland cytology cannot be easily classified. We aimed to design a new algorithm that could differentiate between these neoplasms based on a scoring system. We proposed a new scoring system that enables to reproducibly diagnose non-invasive papillary urothelial tumors. In this system, each lesion was given individual scores from 0 to 3 for mitosis and cellular thickness, from 0 to 2 for cellular atypia, and an additional score for papillary fusion. These scores were combined to form a summed score allowing the tumors to be ranked as follows: 0-1 = UP, 2-4 = low malignant potential (LMP), 5-7 = low-grade transitional cell carcinoma (TCC), and 8-9 = high-grade TCC. In addition to the scoring system, ancillary studies of MIB and p53 indexes with CK20 expression pattern analyses were compared together with clinical parameters. The MIB index was strongly correlated with disease progression. Four of the 22 LMP patients (18.2%) had late recurrences, two of these four (9.1%) had progression to low-grade carcinoma. The MIB index for LMP patients was strongly associated with recurrence (recurrence vs. non-recurrence, 16.5 vs. 8.1, p < 0.001). The proposed scoring system could enhance the reproducibility to distinguish papillary urothelial neoplasms.