Background/aim: We aimed to define the optimal SUVmax cut-off value in determination of mediastinal-hilar lymph node metastasis, by comparing positive PET/CT results with the results of endobronchial ultrasound guided transbronchial needle aspiration biopsy (EBUS-TBNA).Materials and methods: Thirty-one patients with malignancy whose PET/CT imaging revealed a hypermetabolic mediastinal and/ or hilar lymph node and who had undergone EBUS-TBNA were evaluated retrospectively. Histopathology was regarded as the gold standard. The diagnostic role of PET/CT in mediastinal/hilar lymph node metastasis was investigated and compared with the results of contrast-enhanced CT.Results: When a SUVmax value of 2.5 was used, the sensitivity, positive predictive value (PPV), and diagnostic accuracy of the PET/CT were 100%, 65.4%, and 65.4% respectively. In the ROC analysis, the SUVmax cut-off value with the highest diagnostic accuracy (75%) was calculated as 6.3, and when this value was considered, the sensitivity, specificity, PPV, negative predictive value, and diagnostic accuracy of the PET/CT were determined as 70.6%, 83.3%, 88.9%, 60%, and 75% respectively (AUC: 0.779). The sensitivity, PPV, and diagnostic accuracy of the thorax CT were calculated as 91.1%, 72%, and 71.1%, respectively. Conclusion:When determining mediastinal-hilar lymph node metastasis via PET/CT, although a SUVmax cut-off value of 6.3 increases specificity and diagnostic accuracy, we think that a SUVmax cut-off value of 2.5 and above give more optimal results in routine practice.