Sixteen patients, ages 14 to 18, eleven with isolated gonadotropin deficiency and five with sporadic multiple pituitary hormone deficiency, were subjected to a course of five daily intramuscular injections of synthetic luteinizing hormone releasing hormone (LH-RH), 100 mug/day. Before and after the course of intramuscular injections, a rapid LH-RH test (by a one-bolus intravenous injection of 50 mug/sq m) was performed and the responses of plasma LH and follicle-stimulating hormone were measured by a radioimmunoassay method. The patients could be divided into three groups according to the response of the plasma LH to the second LH-RH test: group A, five patients with a significantly higher response of plasma LH to the second LH-RH test: group B, nine patients with a less significantly higher response of the plasma LH to the second LH-RH test; and group C, two patients with very low or no response to either stimulation used in this study. The patients in the three groups may represent different etiologic entities, namely that of a separate hypothalamic lesion, a "mixed" pituitary and hypothalamic lesion, and a "pure" pituitary lesion, respectively. It is concluded that the proposed procedure provides a useful tool for discriminating etiologic groups in patients with abnormal gonadotropic secretion. Recognition of tertiary hypogonadism (primary, pure, hypothalamic gonadotropin-releasing hormone deficiency) is of practical importance in selecting those patients who can benefit from long standing LH-RH therapy.