The purpose of this study was to evaluate the feasibility of 2-[fluorine-18]fluoro-2-deoxy-d-glucose (FDG) and positron emission tomography (PET) for rapid detection of human infections. Eleven patients who were known or suspected to be harboring various infections were studied with FDG-PET. Dynamic scans over the putative infection sites were performed immediately after FDG (370 MBq) injection through 60 min, and static images including multiple projection images were then obtained. FDG uptake was assessed visually into four grades (0, normal; 1, probably normal; 2, probably abnormal; 3, definitely abnormal). For the semiquantitative index of FDG uptake in infections, the standardized uptake value of FDG normalized to the predicted lean body mass (SUV-lean, SUL) was determined from the images obtained at 50-60 min after FDG injection. PET results were compared with final clinical diagnoses. Eleven lesions in eight patients, which were interpreted as grade 2 or 3 by FDG-PET, were all concordant with active infectious foci. The SUL values of infections ranged from 0.97 to 6.69. In two patients, FDG-PET correctly showed no active infection. In one patient, it was difficult to detect infectious foci by FDG-PET due to substantial normal background uptake of FDG. In total, FDG-PET correctly diagnosed the presence or absence of active infection in 10 of 11 patients. Fusion images of PET with computed tomography showed the most intense FDG uptake to be within an abscess wall. In conclusion, FDG-PET appears to be a promising modality for rapid imaging of active human infections. More extensive clinical evaluation is warranted to determine the accuracy of this method.