BACKGROUND: p16 overexpression is a highly sensitive yet moderately specific biomarker for predicting human papillomavirus (HPV)-associated oropharyngeal squamous cell carcinoma (OPSCC). Nuclear β-catenin translocation has been linked to HPV-positive OPSCC. However, whether the strategy of combining β-catenin with p16 can better predict HPV-associated OPSCC remains unknown. METHODS: We evaluated the expression of p16 and β-catenin (nuclear and membrane) by immunohistochemistry staining in 101 OPSCC tissues and HPV status by HPV DNA in situ hybridization. Logistic regression models were used to evaluate single or multiple biomarkers for HPV prediction. The prediction power, sensitivity, and specificity were determined by receiver operating characteristic (ROC) analyses. RESULTS: Our data showed that upon univariate analysis, p16 and nuclear β-catenin were positively correlated with HPV status, while membrane β-catenin was inversely correlated with HPV status (P < 0.01). p16 showed the highest HPV predictive power, with area under the curve (AUC) of 0.9074 compared to 0.6762 for nuclear β-catenin and 0.7635 for membrane β-catenin, respectively, indicating differential accuracies for HPV prediction. Multivariable analysis showed that p16 was significantly correlated with HPV, while nuclear and membrane β-catenin showed marginal significance. The three-biomarker model was similarly sensitive (98.9% vs. 100%) but more specific (88.9% vs. 81%) than p16 alone, which also showed a good predictive value for overall (P = 0.0002) survival and disease-free (P = 0.0158) survival. CONCLUSION: Our study suggests a novel model of combining p16 and subcellular β-catenin for prediction of HPV-associatred OPSCC, and this finding deserves further validation.