The initial clinical application of PET has demonstrated intriguing aspects in ischaemic cerebrovascular disease and epilepsy. A few observations on migraine have also been published, and studies on extrapyramidal disorders are currently underway at the UCLA and the Hammersmith Hospital. The development of PET centres in the USA and Europe will result in larger clinical studies. The capacity for accurate and reliable quantitation of physiological parameters now opens the door to serial measurements in individual patients. Quantitation in absolute units of flow or consumption of a metabolite means that correlations and comparisons can be made between patients, and between studies in individual patients. This means that the natural history of cerebral disease in pathophysiological terms is open to study and, in turn, the effects of therapeutic interventions can be monitored. In this way, with the formulation of appropriate clinical research questions, significant advances in the understanding of cerebral disease and its treatment can be expected. Though the clinical results from PET studies are as yet scanty or preliminary, sufficient data exists to indicate that PET is the most promising tool for the exploration of functional and metabolic processes in the central nervous system in vivo.