Given the abrupt and time-limited nature of daytime-awake and nocturnal-sleep panic attacks, several chemical and neuroendocrine challenge tests have been employed to investigate the neurobiology of "spontaneous" panic attacks. Previously we demonstrated that panic disorder patients have blunted growth hormone (GH) responses to clonidine, an alpha 2-adrenergic agonist. However, the mechanism of this blunted response and the role of hypothalamic-GH dysfunction, if any, remains unclear. To further delineate the status of hypothalamic-GH function in panic disorder, we review the literature and present original data on the GH responses to a number of different chemical and neuroendocrine challenge paradigms. Although stress-mediated increases in GH are thought to be a common correlate of stress in humans, our findings indicate that panic disorder patients have significantly blunted GH responses to clonidine, yohimbine, growth-hormone releasing factor, and caffeine compared to normal control subjects. A similar trend was noted in the delayed rise in GH after glucose challenge. There was no difference in the rate of abnormal GH responses to thyrotropin-releasing hormone in panic disorder compared to normal control subjects. No drug or neuroendocrine challenge, even if associated with marked increases in anxiety, produced a significantly enhanced GH response compared to normal control subjects. These findings provide support for a hyporesponsive hypothalamic-GH system in panic disorder. These observations, combined with preliminary observations from our clinic of short stature in several cases of prepubescent children with anxiety disorders, also underscore the need for assessing early growth patterns in individuals with panic disorder. Strategies for investigating the site(s) of possible neurotransmitter or hypothalamic-GH-somatomedin dysfunction are discussed.