Objectives: To evaluate the survival benefits that lymph node dissection (LND) brought to clinically node-negative upper tract urothelial carcinoma (UTUC) patients. Methods: Non-metastatic node-negative UTUC patients were identified from the Surveillance, Epidemiology and End Results database. N0 patients were naturally divided as cN0-pNx group (clinically diagnosed as N0 without LND performed) and cNx-pN0 group (pathologically diagnosed as node-negative no matter what clinical node status they have). Results: Of the 2731 patients included, 2240 and 491 cases were cN0-pNx and cNx-pN0, respectively. The overall survival (OS) of cNx-pN0 patients was significantly better than that of cNx-pN0 patients (p = 0.022). After propensity score matching, the survival of cNx-pN0 patients was still significantly better than cN0-pNx group. Besides, multivariate analyses showed cNx-pN0 (received LND) was an independent favorable prognostic factor for OS and CSS compared with cN0-pNx (no LND). Survival advantages of pN0 group were more significant in ≥ T2 patients and patients with tumor size ≤ 5 cm. Even in N0 patients who received adjuvant treatment, LND still brought obvious survival improvement (HRos = 0.565, p = 0.013; HRcss = 0.607, p = 0.046). Conclusion: LND could improve survival outcomes in patients with clinically node-negative UTUC, especially for those with muscle-invasive diseases (T2–4 stages) or smaller tumor size (≤ 5 cm). Adjuvant treatment after nephroureterectomy is incapable of replacing the therapeutic role of LND.