In adulthood the growth hormone (GH) response to growth hormone-releasing hormone (GHRH) is inhibited by previous acute administration of either GH or GHRH and it is restored by substances that inhibit hypothalamic somatostatin release. Because in children the GH response to GHRH is not affected by previous neurohormone administration, it has been suggested that in childhood a GH increase is not able to trigger the somatostatin-mediated negative GH autofeedback mechanism. To verify this hypothesis, in 25 children (8 girls and 17 boys; 15 prepubertal and 10 in pubertal stages II-IV) with familial short stature (normal height velocity and insulin-like growth factor I levels) we studied the effect of acute i.v. administration of different recombinant human GH doses (group 1, N = 5, 0.06 U/kg; group 2, N = 6, 0.01 U/kg; group 3, N = 7, 0.005 U/kg at - 150 min) or saline on the GH response to GHRH (1 microgram/kg i.v. at 0 min). In another group (N = 7), we studied the effect of 0.005 U/kg iv recombinant human GH or saline on the GH response to GHRH combined with arginine (0.5 g/kg i.v. over 30 min), which likely inhibits hypothalamic somatostatin release. Serum GH increases after recombinant human GH were dose-dependent (GH peak, mean +/- SEM, 171.7 +/- 24.4, 33.3 +/- 3.9 and 21.8 +/- 5.1 micrograms/l, respectively). The administration of recombinant human GH strongly inhibited the GHRH-induced GH rise in all groups (group 1, 7.1 +/- 1.7 vs 23.1 +/- 7.6 micrograms/l, p < 0.05; group 2, 9.5 +/- 2.8 vs 26.9 +/- 8.5 micrograms/l, p < 0.05; group 3, 9.1 +/- 2.7 vs 34.8 +/- 7.2 micrograms/l, p < 0.02). The GH response to arginine + GHRH (56.9 +/- 13.3 micrograms/l) was higher than that to GHRH alone recorded in group 1 (p < 0.005), group 2 (p < 0.01) and group 3 (p < 0.01), while exogenous recombinant human GH failed to inhibit it (45.0 +/- 9.4 micrograms/l). Our results demonstrate that in childhood, as well as in adulthood, recombinant human GH administration inhibits the somatotrope responsiveness to GHRH. This inhibitory effect is likely to be mediated by hypothalamic somatostatin release.