Human papillomavirus–related oropharyngeal squamous cell carcinoma (OPSCC) has favorable prognosis relative to other head and neck squamous cell carcinomas. Criteria for predicting human papillomavirus status based upon p16 staining, including difficult cases with partial staining patterns, have been developed; however, clinical validation of these criteria and the clinical significance of partial p16 staining have not been reported. Eighty-one archival OPSCC cases were initially stained for p16 by immunohistochemistry with clone G175-405. The percentage of p16+ cells and percentage of confluence of p16+ cells were categorized as 25%, 26% to 75%, or >75%. Of all cases, 16 (20%) had partial p16 expression, with 26% to 75% p16+ cells. Applying previously developed criteria of >75% p16+ cells or >50% positive cells with >25% confluence, 48 (59%) patients were categorized p16+ and demonstrated expected clinical characteristics and superior disease-free survival and overall survival (P<0.001) compared with p16− patients. By themselves, the partial staining patients had intermediate outcomes; however, separating the partial staining cases by degree of confluence showed that those with >75% confluence had superior disease-free survival (P=0.042). When the 16 original partial staining cases were re-stained with the alternative anti-p16 E6H4 clone, p16 status remained concordant for all cases, but only 3 of the 16 were interpreted as demonstrating partial staining. This report shows that the prevalence of partial p16 staining varies with the antibody utilized and clinically validates the application of a graded evaluation of both the number as well as confluence of positive cells for risk stratification of patients with OPSCC.