The aim of this study is to assess the different metabolic activities characteristic of glioma recurrence and radiation necrosis (RN) and to explore the diagnostic accuracy for differentiation of the two conditions using (11)C-methionine (MET), (11)C-choline (CHO), and (18)F-fluorodeoxyglucose (FDG)-positron emission tomography (PET). Fifty patients with lesions suggestive of recurrent glioma by MRI underwent MET, CHO, and FDG-PET. All patients who had previously been treated with radiotherapy for malignant glioma were subjected to open surgery and pathological diagnosis (17 recurrent grade 3- gliomas (Gr.3s) comprising 7 anaplastic astrocytomas (AAs) and 10 anaplastic oligodendrogliomas (AOs), 17 recurrent glioblastomas (Gr.4s), and 16 RNs). We measured the PET/Gd volume ratio, the PET/Gd overlap ratio, and the lesion/normal brain uptake ratio (L/N ratio) and determined the optimal index of each PET scan. The PET/Gd volume ratio and the PET/Gd overlap ratio for RN were significantly lower than those of glioma recurrence only with MET-PET (P < 0.05). The L/N ratio of RN was significantly lower than that of Gr.4 with all PET imaging (P < 0.001) and was significantly lower than that of Gr.3, especially for AO, only with MET-PET images (P < 0.005). Receiver operating characteristic (ROC) analysis showed that the area under the curve of MET, CHO, and FDG was 92.5, 81.4, and 77.4, respectively. MET L/N ratio of greater than 2.51 provided the best sensitivity and specificity for establishing glioma recurrence (91.2% and 87.5%, respectively). These results demonstrated that MET-PET was superior to both CHO and FDG-PET for diagnostic accuracy in distinguishing glioma recurrence from RN.