The role and extent of neck dissection in primary parotid cancer are controversial. Herein, we characterize patterns of lymph node metastasis in parotid cancer. Retrospective analysis. National Cancer Database. Patients with the 6 most common histologic subtypes of parotid cancer were selected. Primary outcomes were the distribution of positive lymph nodes by level and overall survival assessed by Cox analysis. Secondary outcomes included predictors of extended lymph node involvement (≥3 lymph nodes or Level IV/V involvement), via logistic regression. Six thousand nine hundred seventy-sevenpatients with acinic cell carcinoma, adenocarcinoma, adenoid cystic carcinoma, carcinoma ex pleomorphic adenoma (CExPA), mucoepidermoid carcinoma, and salivary duct carcinoma (SDC) were included. Among cN0 patients, 8.2% of low-grade tumor patients had occult nodal metastasis versus 30.9% in high-grade tumor patients. Elective neck dissection was not associated with an overall survival benefit (adjusted hazard ratio: 1.10; 0.94-1.30, p = .238). Among cN+ tumors, CExPA (odds ratio [OR]: 1.88, 1.05-3.39, p = .034) and high-grade pathology (OR: 3.03, 1.87-4.93, p < .001) were predictive of having ≥3 pathologic nodes. CExPA (OR: 2.13, 1.22-3.72, p = .008), adenocarcinoma (OR: 1.60, 1.11-2.31, p = .013), SDC (OR: 1.92, 1.17-3.14, p < .01), and high-grade pathology (OR: 3.61, 2.19-5.97, p < .001) were predictive of Level IV/V neck involvement. In parotid malignancy, nodal metastasis distribution is dependent on histology and grade. High-grade tumors and certain histologies (SDC and adenocarcinoma) had a higher incidence of occult nodes. Comprehensive neck dissection should also be considered for node-positive high-grade tumors, SDC, and adenocarcinoma.