In this study, an examination was made of the sites in the brain of the rat at which the injection of cholecystokinin octapeptide (CCK-8) would alter food intake. Rats fasted for 24 hr received intracerebral injections of CCK-8 (1 nmol) or an equal volume of saline (0.5 microliters), into various sites in the brain through permanently implanted stainless steel cannulae. After prior acclimatisation to individual plexiglass compartments, latency to feed, as well as consumption of food and water during 0-20, 20-40 and 40-60 min after the injection, were recorded. The available food was the standard rat pellets, to which the animal otherwise had constant daily access. With this paradigm, active sites at which CCK-8 suppressed feeding were defined as sites at which consumption of food for 0-20 min was reduced by 25% or more, or the latency to feed was increased by 3 min or more after the injection of CCK-8, as compared to the effect of the injection of saline, made at the same site. Such active sites were most densely distributed in the rostral diencephalon, e.g. hypothalamus, the medial pontine area and lateral medulla, in the vicinity of the nucleus tractus solitarii (NTS). By grouping data for injections according to histologically identified sites, statistical analysis of groups of injections confirmed that these three major areas of the brain were active with regard to the suppression of feeding by CCK-8. These data suggest that CCK may not only initiate satiety messages, as a circulating hormone at peripheral sites, but also participate in the conduction of such information to the target in the brain by serving as a neurotransmitter in the lateral medulla (e.g. NTS), medial pontine area (e.g. relay station between the NTS and hypothalamus) and the lateral hypothalamus, where local release of CCK-8 after stomach loading has been observed.