PurposeAlthough kidney-sparing surgery (KSS) is a nonminor option for low-risk upper urinary tract urothelial cancer (UTUC), its oncological benefits in high-risk UTUC remain unclear when compared to radical nephroureterectomy (RNU). This study aimed to compare the oncological outcomes of RNU and KSS in patients with UTUC. MethodsWe searched the SEER database for patients treated for primary non-metastatic UTUC with either RNU or a kidney-sparing approach (segmental ureterectomy (SU) or local tumor excision (LTE)) between 2004 and 2018. ResultsThe study included 6,659 patients with primary non-metastatic UTUC treated with surgery; 2,888 (43.4%) and 3,771 (56.6%) patients presented with ureteral and renal pelvicalyceal tumors, respectively. Finally, 5,479 (82.3%) patients underwent RNU, 799 (12.0%) were treated with SU, and 381 (5.7%) patients received LTE. For confounder control, propensity score matching (PSM) of patients treated with SU and RNU was performed to adjust for T stage, grade, age, gender, tumor size, and lymphadenectomy performance. PSM analysis included 694 patients treated with RNU and 694 individuals who underwent SU. In multivariable Cox regression and Kaplan-Meier analyses, we found no difference in either CSS or OS between RNU and SU, even in the subgroup of high-grade and/or muscle-invasive UTUC including pT3-T4 tumors (all p > 0.05). ConclusionIn this population-based study, SU provides equivalent CSS and OS compared to RNU, even in high-risk and locally advanced ureteral cancer. Due to the unavoidable risk of selection bias, further prospective studies are expected to overcome the limitations of this study and support the wider implementation of KSS.