Positron emission tomography (PET) has enabled us to study the human brain with unrivalled sensitivity, and has already established its place in the research of neurological conditions such as Parkinson's disease and epilepsy. PET has been used as a tool in the study of patients with motor neuron disease (MND) for well over ten years now, but its potential in diagnosis and to identify surrogate markers of disease expression (phenotype) and progression has yet to be fully realized. The early studies using 2-18fluoro-2-deoxy-D-glucose to measure regional changes in cerebral metabolic rate for glucose gave the first clues to the more widespread involvement of the brain in MND. Later studies exploited the development of activation studies using 15O-containing tracers, which allowed correlation with neuropsychological measures, and the refinement of mapping techniques to delineate the extra-motor areas involved in the disease process. More recently, studies involving ligands such as 11C-flumazenil have allowed the exploration of functional reorganisation in MND, and inhibitory interneuronal pathways which may be crucial in modulation of disease expression. In the future new ligands will be applied in combination with other modalities of investigation (multimodal magnetic resonance imaging; neurophysiological studies) in order to understand the pathophysiology of this heterogeneous condition. Although the potential of PET has not yet been realized in ALS, it is likely to play a part in defining new diagnostic and surrogate markers of disease extent and severity.