OBJECTIVE The purpose of this retrospective study was to determine the sensitivity and specificity of combined PET/CT in differentiating benign from malignant adrenal nodules measuring at least 1 cm in diameter in patients with cancer. MATERIALS AND METHODS We reviewed the radiology reports and images of patients with known malignant disease who had undergone PET/CT for cancer staging or surveillance and who had adrenal nodules at least 1 cm in diameter. We identified 112 adrenal nodules in 96 patients. Two-dimensional PET had been performed 1 hour after administration of (18)F-FDG. Unenhanced CT was performed for attenuation correction, to determine lesion size, and for coregistration with PET data. Adrenal nodules were considered to have a positive PET result if the average standardized uptake value was greater than that of the liver. Follow-up data and biopsy reports were used to determine the pathologic status of the adrenal nodules. RESULTS Thirty adrenal lesions were malignant. Twenty-five of the 30 malignant nodules had positive PET results. Twelve of 82 benign nodules were PET positive with a sensitivity of 83.3% and specificity of 85.4%. Patients with four of five malignant nodules with negative PET results had received previous therapy. The positive predictive value for detection of malignant lesions was 67%, and the negative predictive value was 93%. CONCLUSION Adrenal masses that are not FDG avid are likely to be benign with a high negative predictive value. Especially in patients undergoing therapy, however, there is a small but statistically significant false-negative rate. A considerable proportion of benign nodules have increased FDG activity.