This article presents a series of experiments that involves the development of three novel strategies for human stress research and the utilization of these strategies to examine neurobehavioral processes of stress in healthy volunteers, schizophrenia, and affective illness. The first strategy involved intravenous 2-deoxy-D-glucose (2DG) administration, a glucoprivic stressor. We found that glucoprivic stress results in dissociation of hypothalamus-pituitary-adrenal (HPA), adrenomedullary, and sympathoneural activity. In addition, glucoprivic stress in neuroleptic-treated schizophrenic patients caused heightened dopamine activity, as reflected by increased plasma homovanillic acid (HVA) levels and decreased adaptive responses as assessed by decreased food consumption following 2DG administration. These data suggest that neuroleptics do not prevent stress-related increases in dopamine activity and that schizophrenia may be associated with abnormalities in the stress response. The second strategy assessed effects of uncontrollable and identical amounts of controllable stress in volunteers and depressed patients. In volunteers, it was found that uncontrollable in comparison to controllable stress results in specific behavioral and neuroendocrine alterations. Moreover, uncontrollable stress exposure in depressed patients in comparison to volunteers produced greater alterations in behavioral ratings and plasma cortisol levels and that the uncontrollable stress related increases in helplessness ratings and cortisol levels were significantly correlated. These data suggest that depressed patients may have increased sensitivity to uncontrollable stress and that there may be an important interrelationship between the cognitive deficits of depression and the heightened HPA axis activity observed in these patients. Lastly, we used a naturalistic strategy to examine mechanisms relating childhood parental loss and the development of adult affective illness and found that among subjects with early parental loss histories, those who developed adult psychiatric illness had increased resting plasma levels of cortisol and beta-endorphin (ir) as compared with subjects with early loss and no adult history of psychiatric illness. Moreover, increased HPA activity in adulthood was significantly related to poor childhood adjustment to parental loss. The implications of the results of these studies are discussed.