As part of a prospective biomarker study, we tested oropharyngeal cancer tissue with a highly sensitive in vitro reverse hybridization assay for human papilloma virus (HPV). We tested the sensitivity and specificity of clinical HPV and p16 assays. We also hypothesized that HPV subtypes may be associated with clinical outcomes. From May 2014 through October 2017, approximately 502 participants were screened at Moffitt Cancer Center as part of an ongoing prospective biomarker study. Eligible patients were 18 or older, male, with newly diagnosed treatment naïve squamous cancer of the oropharynx. After exclusions, 204 participants enrolled and were eligible, and 161 had both tumor HPV subtyping data and clinical data for analysis. Tumor samples were tested using in vitro reverse hybridization assay RHA Kit HPV LiPA25 for the qualitative identification of HPV DNA, including 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 68, 6, 11, 44, 53, 66, 70, and 74. Patients were classified according to HPV genotype, HPV 16 only (HPV16+) subtype, HPV 16 in combination with other (HPV16+other) subtypes, HPV positive non-16 (HPVnon16+) subtypes, or HPV negative. Locoregional control (LRC), freedom from distant metastasis (FFDM), and overall survival (OS) based on HPV genotype was estimated according to Kaplan-Meier method and comparisons were made by log rank test. HPV subtyping demonstrated that 78.3% (n=126) were solely HPV16+, 5.6% (n=9) were both HPV16+other, 8.7% (n=14) were HPVnon16+, and 7.5% (n=12) were HPV-. Interestingly, while only 66 patients had their HPV clinically tested, there were significant discrepancies, with a sensitivity of the clinical tests of 71.7%, and a specificity of 33.3%. Immunohistochemistry with p16 (n=159) performed better with a higher sensitivity of 91.8%, and a specificity of 33.3%. There were no differences in patient sex, age, stage, smoking status, or treatment modalities among the different HPV subtype groups. At a median follow up time of 20 months, the 2-year LRC for HPV16+, HPV16 +other, and HPVnon16+ was 92%, 88.9%, and 75.2% respectively (p=0.026 for the comparison of HPV16+ and HPVnon16+). There were no significant differences in 2-year FFDM by subtype: 81.1% (HPV16+), 88.9% (HPV16+other), and 84.6% (HPVnon16+). Actuarial rates of 2-year OS were not significantly different: 81% for HPV16+, 100% for HPV16+other, and 68.2% for HPVnon16+ (p=0.239 for HPV16+ vs HPVnon16+ comparison). In this prospective biomarker study, it appeared that clinical HPV testing may not be as reliable as p16 immunohistochemistry. Patients with HPV16+ detected in their tumors appear to have a better prognosis than those with only HPVnon16+. Further follow-up is necessary, but these data may have implications in patient selection for future staging or de-escalation strategies.