It has been well established that hepatocellular carcinomas (HCCs) accumulate [18F]fluorodeoxyglucose (FDG) to varying degrees; this is thought to be due to differing amounts of FDG-6-phosphatase activity. The purpose of this study was to evaluate the impact of FDG imaging on the management of patients diagnosed with hepatocellular carcinoma. We conducted a retrospective review of the clinical data of 91 consecutive patients diagnosed with HCC who underwent FDG-positron emission tomography (PET) imaging between August 1993 and March 2001. The patients were divided into two groups. In Group one 67 of 91 (74%) patients were evaluated for proven but untreated hepatic lesions using PET. In Group two the remaining 24 patients (26%) were referred for evaluation of HCC recurrence but did not have prior PET. The FDG images were acquired with two dedicated PET tomographs [Siemens ECAT 933, CTI (Knoxville, TN) and GE Advance, General Electric Medical Systems (Milwaukee, WI)] one hour after the intravenous administration of 10 mCi of FDG. Tumor biopsy or resection specimens were available for review from 34 patients and were evaluated for histologic grade, presence of cirrhosis, tumor necrosis, and intratumoral fibrosis. In group one 43 of 67 (64%) of the HCCs accumulated FDG. Sixteen of the 43 patients in whom FDG was accumulated had multiple subsequent FDG-PET scans either for monitoring therapy or for detection of recurrence. FDG-PET imaging had an impact on the management of 20 of these patients: by guiding the biopsy at the metabolically active site of a large necrotic tumor (one), by identifying distant metastases (five), by monitoring the response to treatment with hepatic chemoembolization and guiding additional regional therapy (12), and by detecting recurrence (two). In group two recurrence and/or metastases were demonstrated with FDG-PET imaging in six of 24 (25%) patients, three of whom had multiple subsequent FDG-PET scans to monitor their treatment. Higher histopathologic grade and intratumoral fibrosis but not necrosis or cirrhosis correlated with PET positivity. In this study only 64 per cent of HCCs accumulated FDG. Despite this limitation FDG-PET imaging remains a useful tool in the diagnosis and treatment of HCC. FDG-PET imaging had a clinically significant impact in 26 of 91 (28%) patients with HCC. This includes detection of unsuspected metastatic disease in high-risk patients-including liver transplant candidates-and monitoring response to hepatic-directed therapy. FDG-PET should be considered as part of the workup and management of selected patients with HCC.