Ghrelin was purified from rat stomach as an endogenous ligand for the GH secretagogue (GHS) receptor. As a GHS, ghrelin stimulates GH release, but it also has additional activities, including stimulation of appetite and weight gain. Plasma GH and ghrelin secretory patterns appear unrelated, whereas many studies have correlated ghrelin variations with food intake episodes. To evaluate the role of endogenous ghrelin, GH secretion and food intake were monitored in male rats infused sc (6 mug/h during 10 h) or intracerebroventricularly (5 microg/h during 48 h) with BIM-28163, a full competitive antagonist of the GHS-R1a receptor. Subcutaneous BIM-28163 infusion significantly decreased GH area under the curve during a 6-h sampling period by 54% and peak amplitude by 46%. Twelve hours after the end of treatment these parameters returned to normal. Central treatment was similarly effective (-37 and -42% for area under the curve and -44 and -49% for peak amplitude on the first and second days of infusion, respectively). Neither peripheral nor central BIM-28163 injection modified GH peak number, GH nadir, or IGF-I levels. In this protocol, food intake is not strongly modified and water intake is unchanged. Subcutaneous infusion of BIM-28163 did not change plasma leptin and insulin levels evaluated at 1200 and 1600 h. On the contrary, central BIM-28163 infusion slightly increased leptin and significantly increased insulin concentrations. Thus, endogenous ghrelin, through GHS-R1a, acts as a strong endogenous amplifier of spontaneous GH peak amplitude. The mechanisms by which ghrelin modifies food intake remain to be defined and may involve a novel GHS receptor.