PurposePure ureter cancers are rare and account for only 1–3 % of urothelial carcinomas with limited data. Nowadays, nephron-sparing methods are reserved mainly for imperative cases. This study intends to assess the oncologic outcome between segmental ureterectomy (SU) and radical nephroureterectomy (RNU) for pure ureteral urothelial carcinoma.MethodsFrom July 2004 to August 2010, 112 patients at a single tertiary referral center were included. Perioperative data were obtained from our institutional database. Postoperative CT scan, cystoscopy, and contralateral renal echo were performed regularly for survey of disease recurrence.ResultsThe mean length of follow-up was 43.8 and 48.3 months for the RNU and SU group, respectively. The bladder recurrences, local recurrences, distant metastasis, and cancer-specific survival rates showed no significant differences between RNU and SU (36.4 vs. 34.2 %, p = 0.83; 23.4 vs. 14.3 %, p = 0.27; and 16.9 vs. 8.6 %, p = 0.244, and 13.0 vs. 5.7 %, p = 0.249, respectively).ConclusionThe study suggested that SU is not inferior to RNU for ureter cancer in oncologic outcomes and is less invasive and better nephron preservation.