Gamma-hydroxybutyric acid (GHB), a physiological metabolite of gamma-aminobutyric acid (GABA) was administered to normal persons and patients with hypothalamo-pituitary disorders to examine the effect of GHB on pituitary hormone secretion. GHB stimulated the releases of growth hormone (GH), prolactin (PRL) and cortisol but not luteinizing hormone (LH), follicle stimulating hormone (FSH) and thyrotropin (TSH) in healthy male volunteers. In patients with hypothalamic disorders, plasma GH, PRL and cortisol levels did not increase after GHB administration. In patients with diabetes insipidus, plasma GH responses to GHB were lower than those in normal control, but plasma cortisol and PRL responses were normal. In patients with hypopituitarism, plasma GH and cortisol levels did not elevate after GHB injection, but plasma PRL responses to GHB was normal in three out of patients. In patients with pituitary dwarfism, plasma PRL and cortisol levels increased after GHB injection but plasma GH did not respond to GHB. In patients with acromegaly, plasma GH responses after GHB administration was not consistent. In cases with Cushing's syndrome, plasma cortisol levels increased in patients with adrenocortical hyperplasia but not in those with adrenocortical adenoma. Although plasma GH responses after GHB injection disappeared in patients with Cushing's syndrome, plasma PRL responses was slightly greater than that in normal subjects. In a patient with Addison's disease, plasma ACTH level increased, and plasma PRL and GH responses were normal after GHB injection. To study the relationship between GHB and biogenic amines induced release of human pituitary hormones, the dopaminergic mechanism was blocked by pimozide or stimulated by CB-154 administration, and the serotonergic mechanism was blocked by parachlorophenyl-alanine (PCPA) before GHB administration. GHB induced GH release after treatment with pimozide or PCPA was not significantly different from GH release induced by GHB alone. CB-154 inhibited the increase of plasma GH induced by GHB. Plasma PRL responses to GHB was completely inhibited by CB-154 pretreatment, but not altered by pimozide or PCPA pretreatment. Plasma cortisol responses to GHB was inhibited by pimozide or PCPA pretreatment, but CB-154 had no effect on plasma cortisol responses to GHB injection. GH and cortisol releases induced by GHB was inhibited by continuous TRH infusion. These results suggest that GHB administration is a useful method for examining pituitary hormone secretion. The site of GHB action on pituitary hormone release is probably more central than pituitary in human. Not only inhibition of dopaminergic mechansim but also another mechanism may be involved in GHB induced release of human pituitary hormones.