To investigate the role of central neural cholecystokinin in food intake the concentration of cholecystokinin-like immunoreactivity was measured by radioimmunoassay in the cerebrospinal fluid of male rats. Characterization of the molecular forms of cholecystokinin was made by high-performance liquid chromatography before radioimmunoassay. Four molecular forms of cholecystokinin corresponding to standards of the tetra-, penta-and sulphated octapeptide and a late eluting peak probably corresponding to cholecystokinin-58 were found. The concentration of cholecystokinin-like immunoreactivity in the cerebrospinal fluid decreased in response to 48 h of food deprivation and was restored after 1 h of food intake, the main increase occurring within 30 min after the onset of feeding. Cholecystokinin-like immunoreactivity increased in the cerebrospinal fluid 10 min after an intraperitoneal injection of 5 pg cholecystokinin octapeptide, a dose which also suppressed the amount of food consumed during 1 h in rats deprived of food for 48 h. lntraperitoneal injection of the peripheral, cholecystokinin A receptor antagonists lorglumide (450 pg) or L-364. 718 (20 pg) reversed the inhibitory effect of cholecystokinin octapeptide on food intake and prevented the increase of cholecystokininlike immunoreactivity in the cerebrospinal fluid. It is suggested that central neural cholecystokinin is involved in the control of food intake and that this is reflected in the alterations in cholecystokinin-like immunoreactivity in the cerebrospinal fluid which occur in response to food deprivation and food intake. However, a variety of ways of intracerebral administration of cholecystokinin octapeptide failed to affect food intake in food-deprived rats. The possibility is raised that cholecystokinin octapeptide acts in concert with another transmitter in the brain to affect food intake