Fluorine-18 fluorodeoxyglucose (FDG) positron emission tomography (PET) has proven to be a valuable tool in the initial diagnosis, staging, and restaging of a variety of cancers. The potential use of FDG-PET in the evaluation and management of hepatocellular carcinoma (HCC) continues to evolve. The purpose of this study was to investigate the effectiveness of FDG-PET for the detection and staging of HCC. In addition, we also assessed the correlation between FDG-PET positivity, tumor size, a-fetal protein level (AFP), and histologic grade. All patients on the hepatobiliary and liver transplant service with biopsy proven HCC that underwent FDG-PET between January 2000 and December 2004 were selected for a retrospective review. Results of the FDG-PET scan were compared with other imaging studies [computed tomography (CT), magnetic resonance imaging (MRI), ultrasonography], intraoperative findings, tumor size, AFP levels, and histologic grade. Of the 20 patients who underwent 18F-FDG PET, increased FDG uptake was noted in 14 scans (70%). These 20 patients fell into 2 groups: 1 for detecting HCC (Group A) and 1 for staging HCC (Group B). There were 7 patients in Group A; only 2 scans (28.6%) showed increased uptake. There were 13 patients in Group B; 12 scans (92.3%) showed increased uptake. In Group B, 11 of the 13 scans (84.6%) provided an accurate representation of the disease process. Two scans failed to accurately portray the disease; one scan failed to show any increase in uptake, and the other scan failed to detect positive nodes that were found during surgery. FDG-PET detected only 2 of 8 tumors (25%) < or = 5 cm in size. All 12 PET scans (100%) in tumors > or = 5 cm and/or multiple in number were detected by FDG-PET. FDG-PET scans with AFP levels < 100 ng/ml were positive in 5 of 9 patients (55.6%). In patients with levels > 100 ng/ml, 6 of 7 patients (85.7%) had positive scans. Histologically, there were 6 well-differentiated, 6 moderately differentiated, and 2 poorly differentiated HCCs. FDG-PET detected 4 of 6 for both well- and moderately differentiated HCCs. Both poorly differentiated HCCs were detected. The intensity was evenly distributed between the different histologic grades. There was a strong correlation of FDG uptake with tumor size. There were 5 HCCs with primary tumors >10 cm in size; 4 showed intense uptake on the scan. In contrast, of the 8 tumors < or = 5 cm in size, 6 were negative for uptake. The sensitivity of FDG-PET in detecting HCC < or = 5 cm in size is low and therefore may not be helpful in detecting all of these tumors. For larger tumors, there is a strong correlation of sensitivity and uptake intensity with tumor size and elevated AFP levels. FDG-PET sensitivity and uptake intensity did not correlate with histologic grade. In the setting of extrahepatic disease, FDG-PET seems to be an effective accurate method for HCC staging; however, whether PET offers any benefit over traditional imaging has yet to be determined.