The diagnostic accuracy of 18F-FDG-PET (fluoro-2-deoxyglucose-positron emission tomography) remains questionable for primary hepatocellular carcinoma (HCC) but seems to be more promising for restaging and therapy control. Yet, there are no data on FDG-PET in 131I-lipiodol treatment for primary liver cancer. The aim of this study was to relate baseline FDG-PET findings to histologic data and to assess, for the first time, the role of repetitive FDG-PET imaging for follow-up of 131I-lipiodol treatment. Eighteen (18) patients (16 HCC, 2 cholangiocellular carcinoma; CCC) with 36 treatment courses (up to four per patient) had 35 PET exams, including 18 post-treatment follow-up scans in 10 patients (up to three per patient, one without baseline PET; n = 17). Histopathologic results were available in 15 patients. PET results were retrospectively related to histopathologic type, grading, presence of cirrhosis, and tumor size at baseline and compared with computed tomography (CT) during follow-up. Prior to 131I-lipiodol treatment, 8 patients were PET positive and 9 PET negative. Most of the large HCCs were PET positive and most small tumors PET negative (p < 0.05), despite an overlap below 11 cm. There was no identifiable correlation between PET results and degree of tumor differentiation. Overall, 9 of 10 patients with 17 of 18 follow-up scans showed concordant results with CT. The one discrepant case became PET negative after the first treatment course, despite CT-proven tumor growth (false negative). Patient management was not changed due to PET results. In conclusion, large HCCs were significantly more often PET positive, but there was no correlation with the degree of differentiation. Follow-up PET may be useful if the tumor is first demonstrated to be FDG positive.