A study was conducted to compare the effects of arginine infusion and sc glucagon injection on release of growth hormone. Twentyeight subjects received both agents on separate days and served as their own controls. Eighteen of these were tested after 48- and 72-hr fasts in order to separate any direct growth hormone stimulating effect of glucagon from the indirect response due to falling plasma glucose levels following glucagon induced hyperglycemia. Arginine was an effective agent in the majority of patients studied. There was negligible rise of plasma growth hormone in the absence of variation in plasma glucose levels after subcutaneous glucagon injection. That this failure of output of growth hormone was not due to inability to secrete it is confirmed by positive arginine induced growth hormone responses in the same individuals. Our data shows that arginine is a consistent and potent stimulus for growth hormone release, and is superior to glucagon as a test for growth hormone reserve. The occasional subject who does not respond to arginine may exhibit a small positive response to glucagon. Lack of response to arginine or any other stimulus does not necessarily mean a patient has hypopituitarism; an alternate test for growth hormone reserve should be used for confirmation.