The aim of this study is to analyze the application value of NBI endoscopy in finding the concealed primary lesions of misdiagnosis of oropharyngeal cancer. The clinical data of patients with missed oropharyngeal cancer treated in the Department of Otolaryngology Head and neck surgery, the Second Affiliated Hospital of Xi'an Jiaotong University from May 2018 to June 2021, were retrospectively studied, and the missed diagnosis was also analyzed combined with results of NBI endoscopy. In 31 cases of misdiagnosis of oropharyngeal cancer patients, including 25 males and 6 females, there was no significant difference in age, BMI index, course of disease and TNM stage （> 0.05）, and the pharyngeal or cervical symptoms were the first clinical manifestations of them, containing pharyngeal pain in 17 cases（54.8%） , pharyngeal foreign body sensation in 4 cases（12.9%） and unilateral cervical mass in 10 cases （32.3%）. No laryngoscopy was performed （21 cases） or no primary lesion was found by laryngoscopy （10 cases） at initial diagnosis. Among them, "inflammatory lesions" were given anti-inflammatory treatment with ineffective results or surgical resection was explored for suspicious lesions （17 cases）, or imaging examination （9 cases, including 6 cases with CT and MRI, 3 cases with PET-CT） and cervical lymph node biopsy （5 cases） were carried out for further diagnosis. According to these results, they were given ordinary laryngoscope （2 cases） or NBI endoscopy （29 cases） subsequently, finally they were confirmed as oropharyngeal squamous cellcarcinoma after localized biopsy at the suspicious lesions, indicating that the accuracy of NBI endoscopy in finding the concealed primary lesions of oropharyngeal cancer （93.55%） is significantly higher than that of ordinary electronic laryngoscope （6.45%）（²=43.613, <0.01）. NBI endoscopy has unique advantages in finding oropharyngeal cancer in concealed parts such as tonsil, root of tongue, soft palate and lateral wall of oropharynx, which could reduce misdiagnosis of oropharyngeal cancer.