Continuous infusion or pulsatile administration of growth hormone releasing factor leads to decreasing GH levels and GH responses in normal subjects. We have given 50 micrograms GRF 1-44 i.v. four times in 2-hourly intervals to five normal subjects. After 1 week the same protocol was repeated after s.c. administration of 50 micrograms of the synthetic octapeptide somatostatin (SMS 201-995). The GH response to the same GRF doses was higher after the initial GRF pulse and blunted to the following GRF pulses (pulse I: 37:0 +/- 11.2 ng/ml; Pulse II: 5.3 +/- 1.2 ng/ml; pulse III: 5.9 +/- 2.5 ng/ml; pulse IV: 5.9 +/- 3.2 ng/ml; mean +/- SE). When SMS 201-995 was given 60 min before pulsatile GRF administration, the GH secretion pattern was reversed (pulse I: 2.4 +/- 0.7 ng/ml; pulse II: 2.0 +/- 0.9 ng/ml; pulse III: 4.4 +/- 2.1 ng/ml; pulse IV: 11.4 +/- 3.6 ng/ml). Radioimmunoassayable GRF levels were not different before and after administration of SMS 201-995. The half time of disappearance was 8.6 +/- 0.4 min before and 8.0 +/- 0.5 min after SMS 201-995. Basal thyrotrophin and insulin levels, which remained constant over the 8 h period with GRF only, decreased significantly after SMS 201-995 administration. These findings are compatible with a limited releasable GH pool which is exhausted by chronic GRF stimulation but can be conserved by prior administration of the somatostatin analogue. Thus, when somatostatin bioactivity tapers off, there is recovery of GRF-stimulated GH secretion.