PURPOSE: To establish an objective method of [18F]-fluorodeoxyglucose positron emission tomography (FDG-PET) that can assist in the diagnosis of glucose transporter 1 deficiency syndrome (GLUT1-DS). METHODS: FDG-PET was performed in 8 patients with a mean age of 12.5 years (range, 2-22 years) with GLUT1-DS. Their PET findings were compared with those of 45 controls with a mean age of 11.2 years (range, 2-21 years) by statistical parametric mapping (SPM12, Welcome Neurological Institute). The controls had epilepsy of unknown etiology and normal MRI findings. The age-adjusted ratios of mean radioactivities in regions of interest (ROIs) of bilateral lenticular nuclei, thalami, and the whole cerebral cortex were also measured. The sensitivities and specificities of the ratios for the differential diagnosis of GLUT1-DS were also determined. RESULTS: SPM showed significantly decreased uptake in bilateral thalami and increased uptake in bilateral lenticular nuclei in patients with GLUT1-DS. There were no areas in the cerebral cortex with significant differences between patients and controls. On ROI analysis, by setting the cut-off value of the age-adjusted lenticular nuclei/thalami radioactivity ratio to 1.54, patients with GLUT1-DS were differentiated from controls with sensitivity of 1.00 and specificity of 0.98. CONCLUSION: The age-adjusted lenticular nuclei/thalami radioactivity ratio on PET can distinguish patients with GLUT1-DS from patients with epilepsy of unknown etiology with high sensitivity and specificity. It is important to pay attention to the metabolism of the lenticular nuclei and thalami on PET for the diagnosis of GLUT1-DS.