The uraemic syndrome is characterized by several endocrinological disturbances. This study was undertaken in order to evaluate the GH response to growth hormone-releasing hormone (GRH) in children with chronic renal failure (CRF) and to compare the results with those observed after insulin hypoglycaemia. Twenty-two children with CRF, 10 undergoing continuous ambulatory peritoneal dialysis (CAPD) and 12 on conservative treatment (CT), age ranges 2\p=n-\15 years, were studied and the data were compared with those from 14 children with normal renal function and normal hormonal behaviour, affected by short stature (NC), and those form 13 healthy adult volunteers (NA). The GRH test (1 \g=m\g/kgbody weight, iv) was carried out in 8 CAPD, 8 CT, 9 NC and 10 NA subjects. The blood samples were taken every 30 min for 3 h in CAPD and CT and for 2 h in NC and NA starting at 09.00 h. The following hormones were measured: GH, LH, FSH, Prl, TSH and cortisol (F). The insulin test (0.1 U/kg body weight, iv) was carried out in 5 CAPD, 5 CT, 10 NC and 9 NA on blood samples taken every 30 min for 2 h, measuring GH and glycaemia. No adverse effects were observed after the infusion of GRH. GRH administration induced a prompt response in all subjects, but GH plasma levels were significantly higher in uraemic children than in adults (peak value of 43.5 \ m=+-\ 8.2, 45.0 \ m=+-\ 8.4, 27.8 \ m=+-\ 6.0; 13.5 \ m=+-\ 2.6 \g=m\g/ml in CAPD, CT, NC and NA, respectively). The secretory areas were significantly narrower in NC (P < 0.05) and NA (P < 0.01) than in CAPD, and in NA than in CT (P < 0.01). The GH response to insulin did not differ in the 4 groups. The secretory area in CAPD and CT was wider after GRH than after insulin. The GH peak value of CAPD and NC was significantly higher after GRH than following insulin. No significant variation in TSH, LH and FSH was observed after the infusion of the neuropeptide, whereas Prl and F showed a reduction.The behaviour of Prl and F in NA and NC was similar after placebo and GRH.Our data show: a) There is a greater response of GH to GRH in children than in adults; b) compared with the insulin test, GRH stimulation seems to be a reliable means of evaluating GH secretion both in normal and uraemic children owing to the absence of qualitatively different responses and to the freedom from the adverse effects or risks which follow insulin infusion; c) GRH administration does not induce any significant variation on other hypophyseal hormones and the reduction of Prl and F seems to follow the normal sl eep\ x=req-\ wake and circadian behaviour.