Forty children with short stature were evaluated for GH reserve after pharmacological tests and after a single iv injection of human pancreatic tumor GH-releasing hormone [hpGRH-40). These children were grouped into four diagnostic categories: 1) idiopathic GH deficiency (n = 10); 2) organic hypopituitarism (n = 7); 3) intrauterine growth retardation (n = 5); and 4) constitutional delay of growth and/or familial short stature (n = 18), by standard clinical criteria and physiological and pharmacological tests of GH reserve. Venous blood was sampled for GH concentrations on 2 consecutive days: on day 1, after the iv administration of L-arginine (0.5 g/kg for 30 min) and oral administration of L-dopa (9 mg/kg), and on day 2, after the administration of hpGRH-40, 3.3 micrograms/kg, as an iv bolus. No GH-deficient patient in categories 1 or 2 increased his/her circulating GH concentration to more than 7 ng/ml after the arginine-L-dopa test; however, six children had marked GH responses after hpGRH-40 administration. As a group the lowest peak responses (mean +/- SE) to GRH were found in the organic hypopituitary (3.4 +/- 1.1 ng/ml) and in the idiopathic GH deficiency (8.2 +/- 2.4 ng/ml) categories. All children in the intrauterine growth retardation and constitutional delay of growth (controls for the GH-deficient children) responded briskly to hpGRH-40, although there was wide variation of the peak GH levels (5-51 ng/ml). Circulating concentrations of somatomedin-C did not differ in subjects in any category 24 h after hpGRH-40 injection when compared to basal values. These data indicate that hpGRH-40 can be employed to evaluate GH reseve in short children and may be useful in the diagnosis of hypothalamic-pituitary disorders.