Ghrelin has been proposed as a natural ligand of the GH secretagogue receptor(s) (GHS-R), which was an orphan receptor activated by synthetic peptidyl (hexarelin) and non-peptidyl (MK-0677) GHS to strongly release GH in animals and humans. Herein we studied: 1) the binding of 125I-labeled human ghrelin to membranes from human hypothalamus and pituitary gland; 2) the ability of human ghrelin (either octanoylated or desoctanoylated), as well as of some GHS and neuropeptides to compete with the radioligand. The saturation binding analysis showed, in both tissues, the existence of a single class of high-affinity binding sites with limited binding capacity. The Bmax (maximal number of binding sites) values of ghrelin receptors in the hypothalamus were significantly greater (p<0.001) than those detected in the pituitary, whereas the Kd (dissociation constant) values in the two tissues were similar. 125I-ghrelin bound to hypothalamic membranes was displaced by ghrelin, hexarelin, MK-0677, various GHS antagonists (EP-80317, [D-Arg1-D-Phe5-D-Trp7,9-Leu11]-substance P) and some natural (cortistatin-14) and synthetic (vapreotide) SRIH-14 agonists. In contrast, no competition was seen in the presence of GHRH-44, SRIH-14 or desoctanoylated ghrelin, a ghrelin precursor that is devoid of GH-releasing properties. In conclusion, this preliminary study firstly demonstrates that ghrelin needs octanoylation to bind its hypothalamo-pituitary receptors. These receptors are the specific binding sites for GHS and their antagonists, as well as for SRIH analogs (vapreotide and cortistatin- 14), but not for native SRIH.