Fluorine-18 deoxyglucose (FDG) is not a very tumour-specific substance, and its accumulation in benign lesions with increased glucose metabolism may give rise to false-positive results and hence cause FDG positron emission tomography (PET) to display relatively low specificity (frequently below 85%). Correct interpretation of FDG PET studies is predicated upon detailed knowledge of morphological abnormalities, and the importance of the correlation of functional and morphological information, as derived from computed tomography or magnetic resonance imaging, is discussed. It is emphasized that image fusion programs cannot substitute for understanding of functional and morphological methods. The reconstruction of PET cross-sections is considered, and it is concluded that an iterative image reconstruction method is to be favoured, given its advantages in reducing image artefacts and improving quantification of radioactivity concentrations. The differentiation of malignant and benign lesions when using FDG PET is then reviewed; false-positive findings may be obtained, for example, in patients with acute inflammatory lesions, chronic pancreatitis, retroperitoneal fibrosis or salivary gland tumours. It is suggested that these problems may be alleviated by means of multitracer studies, e.g. using carbon-11 labelled aminoisobutyric acid for quantification of A-type amino acid transport. Finally, the effects of radiotherapy and chemotherapy on FDG uptake and the problems that accrue from these effects are reviewed. Both radiotherapy and chemotherapy can cause increased FDG uptake, complicating diagnosis and evaluation. Knowledge of the effects of different treatment procedures on regional FDG metabolism is therefore necessary for correct interpretation of the PET data.