The presence of distant metastases at initial evaluation influences treatment selection, since no effective systemic treatment for disseminated head and neck squamous cell carcinoma (HNSCC) is currently available. The reported sensitivity for the detection of distant metastases by contrast-enhanced (ce)CT and FDG-PET(/CT) differs substantially between studies. We hypothesized that these sensitivity values are highly dependent on the reference standard use, e.g., follow-up term. Therefore, we analyze our results of FDG-PET/CT (including chest ceCT) with long-term follow-up and compare these findings with data from the literature, with particular interest in the different reference standards. Forty-six HNSCC patients with high-risk factors underwent pretreatment screening for distant metastases by FDG-PET/CT (including chest ceCT). In 16 (35%) patients, distant metastases were detected during screening (6 patients) or during a mean follow-up of 39.4 months after screening (10 patients). The sensitivity and negative predictive value were 83.3 and 97.2% when 6 months, 60.0 and 89.9% when 12 months, and 37.5 and 72.2% when 30 months follow-up were used as reference standard, respectively. This is comparable with reported studies with similar reference standards. This critical appraisal on the reference standards used in our and reported studies shows room for improvement for the detection of distant metastases to refrain more patients from unnecessary extensive locoregional treatment for occult metastatic HNSCC.