ABSTRACT Oropharyngeal cancer (OPC) is now the most common malignancy of the head and neck region, and OPC associated with the human papillomavirus (HPV) has overtaken cervical cancer as the most common HPV- associated malignancy in the United States. African American (AA) patients demonstrate inferior OPC oncologic outcomes to their non-AA counterparts even when adjusting for HPV effect. It is not known whether reduced treatment response and survival in AA OPC patients is driven solely by unequitable access to care and other socio-economic variables or whether it is impacted at least partially by the interaction between ancestry and cancer biology. No previous study has addressed this question using an adequate OPC clinical dataset with correlative genomic and transcriptomic data in order to address this translationally important question. Therefore, it is not possible to determine whether OPC treatment response and survival in AA patients is partially linked to differential tumor biology using existing datasets. In this application, we will utilize two unique OPC cohorts enriched for AA patients to test the hypothesis that OPC development in AA patients is accompanied by more aggressive tumor biology facilitated by an immunosuppressive tumor immune microenvironment (TIME). In Aim 1 we will correlate self-reported race and calculated ancestry to intrinsic OPC tumor biology and treatment response. Specifically, we will test the correlation between tumor cell multi-nucleation (a feature of aggressive OPC biology recently validated by our team in a multi-institutional OPC cohort), survival, and differential gene expression in AA OPC patients matched to their non-AA counterparts for T classification, HPV status, and smoking history. In Aim 2 we will correlate self- reported race and calculated ancestry to changes in TIME and treatment response in previously untreated AA OPC patients as well as immune checkpoint inhibitor response in AA OPC patients with recurrent/metastatic disease. For both sets of analyses, we will deploy novel, validated machine learning approaches to analysis of conventional histologic slides designed to facilitate rapid clinical translation across multiple other institutions. Completion of the proposed studies will, for the first time, establish the link between race/ancestry, OPC biological behavior, and treatment response. Successful validation of our hypothesis will provide a unique opportunity for the development of precision oncology approaches to AA OPC management aimed at reversing the disparities noted in cancer specific survival for this understudied patient population.