Purpose: Accurate evaluation of metastatic cervical lymph nodes plays a decisive role in the treatment and prognosis of patients with squamous cell carcinoma of the head and neck. The purpose of this study is to investigate the usefulness of FDG-PET for diagnosis of cervical metastasis in the head and neck cancer by comparing with the conventional imaging study. Materials and Methods: The subjects on this study were 30 patients (24 males and 6 females, aged 39 to 76, mean 57.1) diagnosed as pathologic-proven squamous cell carcinomas of the head and neck. All patients underwent preoperative FDG-PET, CT(n=27) or MRI (n=3). Their medical records were reviewed retrospectively. Using pathologic reports as a golden standard, the results of FDG-PET were compared with conventional imaging study (CT/MRI) in the evaluation of cervical metastasis. Results: Thirty patients had five different primary sites which were tongue (11), supraglottis (10), glottis (6), hypopharynx (2) and tonsil (1). A total of 40 neck dissections were performed unilaterally in 20 patients and bilaterally in 10 patients. Of these, 16 showed pathologically positive for lymph node metastasis. The sensitivity and specificity of FDG-PET for the diagnosis of cervical metastasis was 75% and 100% respectively, compared with conventional imaging of 56.3% and 95.8%, respectively. The difference of sensitivity was not statistically significant (p=0.453). Of 5 cases with small metastatic node (<1cm), 3 were detected on PET detected correctly but none were detected by CT. Conclusion: FDG-PET was more accurate than conventional imaging study in the diagnosis of metastatic lymph nodes in squamous cell carcinomas of the head and neck, especially detection of small metastatic node. FDG-PET might be useful adjunct to conventional image in the preoperative evaluation of head and neck squamous cell carcinoma.