Fluorine-18-fluoro-2-deoxyglucose (FDG) positron emission tomography (PET) was used to image 38 patients with known or suspected malignant processes involving the abdomen or pelvis (including primary [n = 2] and secondary [n = 19] liver tumors, recurrent colon carcinoma [n = 1], lymphoma [n = 15], and cavernous hemangioma [n = 1]). PET results were compared with those from concurrent computed tomographic (CT) studies. Conspicuity of metastatic liver lesions on PET images often exceeded that of lesions on CT scans. Two well-differentiated primary liver tumors failed to show increased FDG uptake. Thirteen of the 15 patients with lymphoma had active disease, as determined with CT or clinical findings; 12 of 13 had increased FDG uptake on PET images. The false-negative result occurred in a patient with a low-grade lymphoma. The recurrent colorectal lesion was well demonstrated by PET, but further study is needed to determine whether it can be used to distinguish recurrent disease from radiation-induced changes. PET may be useful in determining the response to oncologic therapy, but PET does not reliably allow exclusion of low-grade lymphoma or well-differentiated hepatic lesions. Clinical correlation is always required in evaluating the significance of PET findings, as inflammatory conditions can also result in increased FDG uptake.