BACKGROUND: Lymph node ratio (LNR) has been reported to reliably predict cancer-specific survival (CSS) in parotid gland cancer (PGC). Our study was designed to validate the significance of LNR in patients with PGC. METHODS: Patients diagnosed with stage I-IV PGC were enrolled from Surveillance Epidemiology and End Results database (SEER,  = 3529), which is the training group, and Sun Yat-sen University Cancer Center database (SYSUCC,  = 99), the validation group. We used X-tile software to choose the optimal cutoff value of LNR; then, univariable and multivariable analyses were performed, assessing the association between LNR and CSS. RESULTS: The optimal cutoff value of LNR was 0.32 by X-tile based on 3529 patients from SEER. Cox proportional hazard regression analysis revealed better CSS for patients with LNR ≤ 0.32 (adjusted hazard ratio [HR] 1.612, 95% confidence interval [95% CI] 1.286-2.019;  < 0.001) compared with patients with LNR > 0.32 in SEER. In the SYSUCC cohort, patients with LNR ≤ 0.32 also had better CSS over patients with LNR > 0.32 ( < 0.001). In N2 and N3 stage groups, patients with LNR ≤ 0.32 had superior CSS outcomes over those with the LNR > 0.32 group, but this benefit was absent in the N1 stage group. CONCLUSIONS: In conclusion, the lymph node ratio turned out to be an independent prognostic factor for cancer-specific survival of PGC in this study. This valuable information could help clinicians to evaluate the prognosis of PGC and suggest that adequate lymph node dissection is necessary.