In order to verify the true GH-releasing effect of glucagon and to explain the mechanism underlying this effect, we studied the effect of glucagon (GLU, 1 mg) administered either iv or im on both basal and GHRH (1 microgram/kg)-induced GH rise in 48 normal short children and adolescents. Moreover, the in vitro effect of GLU on rat anterior pituitary cells was studied. Intravenous administration of GLU induced no significant GH rise. On the other hand, im GLU administration induced a clear-cut GH increase (mean +/- SE GH peak after GLU vs placebo = 25.7 +/- 3.9 vs 10.1 +/- 3.6 micrograms/L, p < 0.01). Intravenous administration of GLU failed to modify the GHRH-induced GH rise either when coadministered with the neurohormone (35.2 +/- 4.1 vs 34.1 +/- 6.0 micrograms/L) or when given 60 min earlier (20.2 +/- 5.8 vs 21.1 +/- 8.3 micrograms/L). Differently from iv GLU, im GLU strikingly potentiated the GH response to GHRH given 90 min later (57.5 +/- 6.3 vs 24.7 +/- 9.1 micrograms/L, p < 0.01). Mean plasma glucose levels increased 30 min after GLU, administered either iv or im, and returned to basal levels 60 min later. GH secretion from dispersed rat pituitary cells was unaffected by incubation with GLU (10(-10)-10(-4) mol/L). Incubation of the cells with 10(-7) mol/L GHRH induced instead a clear-cut stimulation of GH release. In conclusion, our data demonstrate that glucagon per se has not GH-releasing activity as indicated by its uneffectiveness to release GH in vitro and after intravenous administration.(ABSTRACT TRUNCATED AT 250 WORDS)