The detection of lymph node metastases is a major challenge in oral and oropharyngeal squamous cell carcinoma (OSCC and OPSCC). ⁶⁸Ga-NOTA-AE105 is a novel positron emission tomography (PET) radioligand with high affinity to urokinase-type plasminogen activator receptor (uPAR), a receptor expressed on the surfaces of tumor cells. The aim of this study was to investigate the diagnostic value of uPAR-PET/CT (computerized tomography) in detecting regional metastatic disease in patients with OSCC and OPSCC compared to the current imaging work-up. In this phase II trial, patients with OSCC and OPSCC referred for surgical treatment were prospectively enrolled. Before surgery, ⁶⁸Ga-NOTA-AE105 uPAR-PET/CT was conducted, and SUVmax values were obtained from the primary tumor and the suspected lymph nodes. Histology results from lymph nodes were used as the standard of truth of metastatic disease. The diagnostic values of ⁶⁸Ga-uPAR-PET/CT were compared to conventional routine preoperative imaging results (CT and/or MRI). The uPAR expression in resected primary tumors and metastases was determined by immunohistochemistry and quantified digitally (H-score). A total of 61 patients underwent uPAR-PET/CT. Of the 25 patients with histologically verified lymph node metastases, uPAR-PET/CT correctly identified regional metastatic disease in 14 patients, with a median lymph node metastasis size of 14 mm (range 3–27 mm). A significant correlation was found between SUVmax and the product of the H-score and tumor depth (r = 0.67; p = 0.003). The sensitivity and specificity of uPAR-PET/CT in detecting regional metastatic disease were 56% and 100%, respectively. When added to CT/MRI, uPAR-PET was able to upstage 2/11 (18%) of patients with occult metastases and increase the sensitivity to 64%. The sensitivity and specificity of ⁶⁸Ga-NOTA-AE105 uPAR-PET/CT were equivalent to those of CT/MRI. The significant correlation between SUVmax and uPAR expression verified the target specificity of ⁶⁸Ga-NOTA-AE105. Despite the target specificity, the sensitivity of imaging is too low for nodal staging and it cannot replace neck dissection.