The aim of the study was to evaluate the diagnostic accuracy of Human Papillomavirus (HPV) techniques in oropharyngeal cancer. PubMed, EMBASE, the Cochrane Library and clinicaltrials.org were systematically searched for studies reporting methods of HPV detection. Primary outcomes were sensitivity and specificity of HPV detection. In this case, 27 studies were included (n = 5488, 41.6% HPV+). In this case, 13 studies evaluated HPV detection in tumour tissue, nine studies examined HPV detection in blood samples and five studies evaluated HPV detection in oral samples. Accuracy of HPV detection in tumour tissue was high for all detection methods, with pooled sensitivity ranging from 81.1% (95% CI 71.9–87.8) to 93.1% (95% CI 87.4–96.4) and specificity ranging from 81.1% (95% CI 71.9–87.8) to 94.9% (95% CI 79.1–98.9) depending on detection methods. Overall accuracy of HPV detection in blood samples revealed a sensitivity of 81.4% (95% CI 62.9– 91.9) and a specificity of 94.8% (95% CI 91.4–96.9). In oral samples pooled sensitivity and specificity were lower (77.0% (95% CI 68.8–83.6) and 74.0% (95% CI 58.0–85.4)). In conclusion, we found an overall high accuracy for HPV detection in tumour tissue regardless of the HPV detection method used. HPV detection in blood samples may provide a promising new way of HPV detection.