Because of an estimated 20–30% prevalence of occult lymph node (LN) metastases in patients with head and neck squamous cell carcinoma (HNSCC), neck dissection is often proposed, despite its potential morbidity. In this systematic review and meta-analysis, the diagnostic performance of FDG PET-CT in detecting occult LN metastases was evaluated in patients with clinically negative necks (cN0) and in whom histopathology of a neck dissection specimen served as gold standard. Overall, 16 studies out of 2062 screened on PubMed and EMBASE fulfilled the inclusion criteria (n = 1148 patients). Seven of these sixteen studies were split into two or three studies because they contained data that could be processed distinctly in our meta-analysis. For this reason, a total of 25 studies were identified and included in the analysis (n total = 1918 patients). The overall prevalence of metastatic nodes per patient was 22.67%. The pooled sensitivity, specificity, diagnostic odds ratios, and negative predictive value (NPV) were 0.71 (95%CI: 0.66–0.75), 0.90 (95%CI: 0.84–0.93), 20.03 (95%CI: 13.51–29.70), and 0.92 (95%CI: 0.89–0.95), respectively. The main causes of inter-study heterogeneity included different reference standards (evaluation per patient, per neck side, or per neck level). The current meta-analysis showed that FDG PET-CT has a high specificity and NPV for ruling out nodal involvement in cN0 necks, but a limited sensitivity.