As breast cancer in patients who have undergone augmentation mammoplasty is usually detected late, prognosis is worse than it would otherwise be. Mammography has led to remarkable improvements in the diagnosis of breast cancer, but did not so in cancer arising from augmented breasts. FDG~PET, which uses fluorine-18-fluorodeoxy glucose (FDG), a glucose non-metabolized analogue, is able to detect breast cancer which shows more active rates of glucose metabolism. Cancer in a radioopaque dense breast, such as breast with silicone-or paraffin-augmentation and is not found by mammography, might be detected by FDG-PET. We tried to find out diagnostic efficiency of FDG-PET in detecting breast cancer with augmented breast. Between June 1995 and November 1997, eight patients had a history of breast augmentation with silicone or paraffin. Although most of them were examined at local clinic by physical exam, mammography or ultrasonography before visiting our hospital, definitive diagnosis could not be made. One patient wanted both breasts augmented with paraffin to be removed, and PET evaluated a total of nine lesions with pathological confirmation. As a result, the mammography detected the breast cancer in only one out of three malignancies, and ultrasonography made false positive result in one patient in augmented breast. In contrast, PET scan predicted all of malignancies and five out of six benign lesions precisely. Two of three breast cancers had axillary FDG uptake interpreted as consistent with metastatic involvement, and in one case with cancer of negative axillary lymph node involvement there was no FDG uptake in the axilla, which were correlated to pathologic finding. Although its high cost made the use of PET as a screening test for all augmented breasts not feasible, it would, however, be the best diagnostic choice if other methods failed.