Facial nerve dysfunction is the principal postoperative complication related to parotidectomy. To test the hypothesis that the modified Sunnybrook Facial Grading System (mS-FGS) is superior to the original S-FGS in the assessment of facial nerve function following parotidectomy. Prospective, longitudinal study evaluating patients with primary or metastatic parotid neoplasms undergoing parotidectomy with facial nerve-sparing between 2016 and 2020. The subjects were assessed twice, on the first postoperative day and at the first outpatient evaluation, 20-30 days post-surgery. Facial assessments were performed using the original and modified (plus showing the lower teeth) versions of the Sunnybrook System and documented by pictures and video recordings. Intra- and inter-rater agreements regarding the assessment of the new expression were analyzed. 101 patients were enrolled. In both steps, the results from the mS-FGS were significantly lower (p < 0.001). Subjects with a history of previous parotidectomy and those who underwent neck dissection had more severe facial nerve impairment. The mandibular marginal branch was the most frequently injured, affecting 68.3% of the patients on the first postoperative day and 52.5% on the first outpatient evaluation. Twenty patients (19.8%) presented an exclusive marginal mandibular branch lesion. The inter-rater agreement of the new expression assessment ranged from substantial to almost perfect. The intra-rater agreement was almost perfect (wk = 0.951). The adoption of the Modified Sunnybrook System, which includes evaluation of the mandibular marginal branch, increases the accuracy of post-parotidectomy facial nerve dysfunction appraisal.