No suggested screening test meets all of the criteria set for such a procedure. The minimum incidence of a positive response in normal children detected in a single blood sample after diethylstilbestrol, sleep or exercise is approximately 70%. This is higher than that observed when a single sample is obtained following oral glucose. While both sleep and exercise are physiologic stimuli, the former may be quite inconvenient unless an outpatient facility staffed with appropriate personnel is available. An exercise test employed in the office may well be the best screening procedure for the practicing physician. The optimal criteria for a definitive test of growth hormone function are also not met by any single stimulus. Insulin-induced hypoglycemia, arginine infusion, intramuscular glucagon and oral 1-DOPA are all useful procedures. None alone discriminate completely between the normal and the growth hormone-deficient child. Despite potential hazards, insulin-induced hypoglycemia remains the standard against which other stimuli are judged. Arginine and 1-DOPA appear to be equally effective. The literature contains insufficient data to allow adequate evaluation of intramuscular glucagon alone, and the results of combined propranolol-glucagon stimulation, while promising, require confirmation. Because of an inconstant and/or small magnitude of response leading to results which are difficult to interpret, the use of glucose, pyrogen, vasopressin and ACTH are not adequate tests of growth hormone function. Bovril® is a satisfactory stimulus for those children who will take it. Those factors which modify the growth hormone response must be considered in evaluating the results of stimulation tests. Blunted responses should be interpreted with extreme caution in the obese child. A fasting growth hormone concentration ≥ 7 ng/ml is presumptive evidence of intact growth hormone function regardless of the subsequent response to stimulation. It is essential that patients be euthyroid in order to interpret the results of growth hormone function tests. Physiologic glucocorticoid replacement therapy should not confuse the interpretation of results. Whether or not pretreatment with sex steroids is worthwhile in the routine evaluation of children for suspected growth hormone deficiency is an open question. Although it is agreed that the definitive diagnosis of growth hormone deficiency depends on the demonstration of failure to respond to two stimuli, which two are most satisfactory is not settled. The sequential administration of arginine and insulin on the same day appears to limit significantly the incidence of false-positive laboratory diagnoses of growth hormone deficiency. The significance of intermediate values in response to stimulation remains unclear. Caution should be exercised in assigning a child to the category of partial growth hormone deficiency. This question must be answered ultimately by the response to HGH therapy in the individual patient. Finally, several points should be kept in mind. All of the tests described depend on the detection and quantitation of immunologically active HGH and biological activity is not necessarily associated with the material(s) being measured. Since many of the stimuli used in the evaluation of growth hormone function are clearly pharmacologic, the physiological significance of the response to such stimuli must be interpreted with caution. The best current evidence suggests that all of the stimuli described act through an intact hypothalamus and pituitary. Differentiation between hypothalamic and pituitary sites of defective growth hormone function awaits the availability of growth hormone-releasing factor(s).