The aim of this retrospective study was to compare the diagnostic accuracy between positron emission tomography (PET) and combined PET/computed tomography (PET/CT) in the detection of recurrent head and neck squamous cell carcinoma (HNSCC) and to evaluate the degree of interobserver agreement. Thirty-two patients who had undergone curative treatment for HNSCC and who presented with a suspicion of recurrent local disease were studied with fluoro-2-deoxy-D-glucose (FDG)-PET imaging. All patients had undergone an inconclusive conventional workup (nasofibroscopy, CT scan and/or MRI). PET and PET/CT were reviewed by two nuclear medicine physicians independently. Performances of PET and PET/CT were compared using biopsy and/or clinical follow-up of at least 8 months as gold-standard. ROC curves were employed for statistical analysis. Out of 32 patients, 18 (56%) had a local recurrence. Intraclass correlation coefficients were strong (>90) and statistically significant (P < 0.0001) for the two reviewers in all cases. The sensitivity, specificity and accuracy of PET were found to be 94%, between 36 and 50% and between 69 and 75%, respectively, depending on the consideration of equivocal cases. Results for PET/CT were found to be 94, 57 and 78%. The utility scores of PET and PET/CT were 0.72 and 0.78, respectively. PET/CT could have a direct impact on patient care with the avoidance of 8/14 (57%) unnecessary invasive procedures (panendoscopy under general anaesthesia). Combined PET/CT is more accurate than PET alone for detection of recurrent HNSCC. The findings of this study are reinforced by the strong interobserver agreement in the interpretation of the results.