A dministration of human pancreatic tumor growth hormone (GH) releasing factor (hpGRF[1-40]) as a single injection to normal human subjects stimulates the secretion of GH in a dose-responsive manner. In the present studies, hpGRF(1-40) was infused in a graded stepwise manner over a 6-h period in order to determine whether the GH secretory response would be sustained. Normal adult males received four consecutive 90-min infusions of hpGRF(1-40) at doses of 1, 3.3, 10, and 33 ng/kg per min, preceded and followed by a 90-min saline infusion; and the plasma GH responses were compared with those during a separate control infusion. Plasma GH levels were significantly elevated by each hpGRF(1-40) infusion; and dose responsiveness was evident for the lowest three doses. Mean integrated GH secretory rates for the four doses were 1.95, 3.29, 4.29, and 3.65 times those of the respective control study. Plasma GH responses exhibited considerable variability, frequently decreasing during the latter part of each infusion; and at the highest dose, they decreased continuously beginning shortly after the onset of infusion. Episodic GH secretion occurred in individual subjects during each of the infusion periods. The possible contribution of hypothalamic somatostatin secretion to the diminished GH responsiveness was evaluated by determining plasma thyroid stimulating hormone (TSH) levels during the infusions and the TSH responses to thyrotropin-releasing hormone (500 micrograms i.v.) during a separate hpGRF(1-40) infusion of 2 ng/kg per min. Neither basal nor stimulated TSH levels differed between GRF-infused and control groups. The results indicate that GH secretion is dose responsive to hpGRF(1-40) infusions, though the response to hpGRF(1-40) infusions, though the response is complex. The absence of impaired TSH secretion provides evidence against a mediating role of somatostatin. The explanation for the loss of GH responsiveness remains undetermined but could include GRF-induced receptor down-regulation, a postreceptor effect, or, in spite of our negative results, a somatostatin-mediated inhibition.