The addition of whole body positron emission tomography (PET) to the investigation of patients with newly diagnosed head and neck squamous cell carcinoma (HNSCC) was assessed over a 6-month period. Staging investigations included laryngoscopy, oesophagoscopy, CXR, CT and MRI. In addition, all patients had an extended-field (whole body) FDG-PET scan and were restaged. Standardised Uptake Values (SUV) were used to measure FDG uptake. SUV levels above 5 were considered indicative of the presence of tumour, values below 3 indicative of benign aetiology and values equal to and between 3 and 5 were considered equivocal. Forty-eight consecutive patients with biopsy proven HNSCC were included for study. Three patients presenting with neck disease had unknown primary tumours. Of the remaining 45 patients, CT scan correctly identified 40 of the primary tumours (89%). MRI and PET both identified 41 primary tumours (91%). Thirty-two patients underwent neck dissection. Of these patients 12 had pathologically N0 necks and 20 had positive nodal disease. CT scan and MRI each correctly staged pN0 necks in 10 of 12 patients (83%) whereas PET alone had a lower true negative rate of 8 out of 12 patients (67%). PET correctly staged the N+ necks in 14/20 patients (70%) versus 12/20 (60%) for MRI, and 8/20 (40%) for CT alone. All four patients who were judged to have distant metastases by PET had these metastases deemed negative by other investigation. None of the three imaging modalities was able to identify the tumour site in the three patients with unknown primaries. In conclusion, although PET has got a higher sensitivity in detecting nodal disease, it has only slightly improved the classification of N+ necks. The findings of this study cast doubt on the merit of routine addition of PET to the current investigative protocols for HNSCC patients.