Magnetic resonance imaging (MRI) has been the most useful tool in the anatomical definition of soft tissue sarcoma, although there remains the problem of defining the lesions as benign or malignant. The management of such lesions requires biopsy prior to surgical resection. If the most malignant area could be defined more accurately, then this area could be targeted for biopsy. Fluorodeoxyglucose positron emission tomography (FDG PET) has been found to be useful in identifying malignancy and variations in grade in soft tissue masses. The aim of this study was to assess the use of FDG PET scanning with or without co-registered MRI to indicate the most appropriate biopsy site. Twenty consecutive patients presented with soft tissue masses with clinical signs of malignancy. All patients underwent MRI and FDG PET scanning and the two images were co-registered. A biopsy site that was the most likely to be malignant was defined on the PET scan. All patients underwent an initial biopsy and then complete surgical resection of the mass. The histological results from the mass were compared with those from the biopsy specimen obtained from the site suggested by the PET scan. In malignant masses the biopsy site suggested by the FDG PET scan was found to be representative of the most malignant site on the whole mass histology. Benign lesions had low or no FDG uptake. In no case did the co-registered image add significantly to the appropriate biopsy site. FDG PET can be used to appropriately direct biopsy in soft tissue sarcoma and potentially may lead to computed tomography/MRI directed outpatient biopsy prior to definitive treatment.