Measurement of the regional cerebral blood flow (rCBF) is an important parameter in the evaluation of cerebral function. With positron emission tomography (PET) rCBF has predominantly been quantified using the short-lived radiotracer oxygen-15 labelled water (H 2 15 O) and an adaptation of the Kety one-tissue compartment autoradiographic model. The values attained in putative grey matter, however, are systematically underestimated because of the limited scanner resolution. For this reason we applied a dynamic kinetic two-tissue compartment model including a fast and a slow flow component each with a perfusable tissue fraction. In the fast component rCBF was 2-2.5 times greater than grey matter values using traditional autoradiography in both human and monkey. Visual stimulation in human gave a corrected rCBF increase of approximately 40%. Visual stimulation was also used to indirectly validate carbon-10 labelled carbondioxide ( 10 CO 2 ), a new very short-lived rCBF PET tracer with a half-life of only 19.3 seconds. This allowed an increase in the number of independent PET scans per subject from 12-14 using H 2 15 O to 64 using 10 CO 2 . The experiment demonstrated a maximal activation response in the visual cortex at a 10-15 Hz stimulation frequency. The use of the rCBF PET mapping technique is illustrated by studies of the organization of language and the oculomotor system. With respect to the former, we found confirmation of neuropsychological evidence of the involvement of the left supramarginal/angular gyrus in reading in Japanese of a phonologically based script system, Kana, and of the left posterior inferior temporal gyrus in reading of a morphogram based script system, Kanji. Concerning the organization of the oculomotor system we found overlapping areas in fronto-parietal cortex involved in maintaining visual fixation, and performing visually guided and imagined eye movements. These data show that overt eye movements are not a prerequisite of the activation of classical cortical oculomotor regions and underscore the involvement of these areas in other behaviours such as visual attention and saccade inhibition. During eye movements in the dark an increased activation response in the parieto-occipital cortex can be found. This can be interpreted as effects of the gaze-sensitive neurons that are used to objectively localize objects relative to the body, and efferent copies of motor commands, used to predict the visual consequences of eye movements to maintain visual continuity. Defect efferent copies are in some neurobiological models of schizophrenia thought to contribute to passivity phenomena. The clinical perspective of brain mapping techniques is to preoperatively locate eloquent areas, e.g. motor function, language, and memory, allowing the achievement of optimal neurosurgical resection with the preservation of neurological function.