To evaluate the clinical utility of voided urine fluorescence in situ hybridization (FISH) for predicting the pathology of preoperative "low-risk" upper tract urothelial carcinoma (UTUC). Information of patients preoperatively diagnosed with "low-risk" UTUC receiving radical nephroureterectomy (RNU) between May 2014 and October 2019 were retrospectively collected. All of the patients accepted the FISH test and then were divided into two groups according to the results of FISH. The diagnostic value of FISH was assessed through the receiver operating characteristics (ROC) curve and area under the curve. Logistic regression analysis was applied to examine FISH as a predictive factor of tumor final stage and grade of preoperative "low-risk" UTUC. In total, 129 patients were included. Of them, 70 (54.2%) were marked with positive FISH result. The difference at final pathology in tumor stage and tumor grade between these two groups of FISH (-) and FISH (+) had significantly statistical significance (p<0.001). Regarding to the tumor stage at final pathology, the sensitivity, specificity, positive predictive value and negative predictive value of FISH were 70.7 (58.9-80.3)68.5 (54.3-80.1)75.7 (63.7-84.8) and 62.7 (49.1-74.7), respectively. Regarding to the tumor grade at final pathology, the sensitivity, specificity, positive predictive value and negative predictive value of FISH were 64.7 (53.5-74.6), 65.9 (50.0-79.1), 78.6 (66.8-87.1) and 49.1 (36.5-62.3), respectively. The results of logistic regression analysis indicated that FISH could predict the pathologic characteristics of preoperative "low-risk" UTUC independently. FISH was qualified with relatively high diagnostic estimates for predicting tumor stage and grade of preoperative "low-risk" UTUC, and could be an independent predictive factor in clinical practice. For preoperative "low-risk" UTUC patients but with positive FISH result, choosing nephron-sparing surgery may require special caution.