The observations and research described in this communication derive from a nontraditional view of the immune system. It has become abundantly clear that there are probably no organ systems or homeostatic defense mechanisms that are not, in vivo, subject to the influence of interactions between behavioral and physiological events. The complex mechanisms underlying these interactions and their relationship to health and illness, however, are imperfectly understood. The most imperfectly understood, perhaps, are the interrelationships among brain, behavior, and immune processes. Without attempting to cover all the literature, we have used stress effects and conditioning phenomena as illustrations to point out that behavior can influence immune function. We have also described data indicating that the immune system can receive and respond to neural and endocrine signals. Conversely, behavioral, neural, and endocrine responses seem to be influenced by an activated immune system. Thus, a traditional view of immune function that is confined to cellular interactions occurring within lymphoid tissues is insufficient to account for changes in immunity observed in subhuman animals and man under real world conditions. These data question seriously the notion of an autonomous immune system. Most of the research on the regulation of immune responses has been predicated on the assumption that such regulation is accomplished by the interacting components of the immune system itself, e.g. interactions among helper and suppressor T-lymphocytes, B-cells, and accessory cells that can result in the production of antibody and effector T cells. The immune system is, indeed, capable of considerable self-regulation, and immune responses can be made to take place in vitro. The functions of that component of adaptive processes known as the immune system that are of ultimate concern, however, are those that take place in vivo. There are now compelling reasons to believe that in vivo immunoregularity processes influence and are influenced by the neuroendocrine environment in which such processes actually take place--an environment that, on the one hand, can generate signals that resting and/or activated leukocytes can receive, and, on the other hand, is exquisitely sensitive to the individual's perception of and capacity to adapt to the demands of the environment. The immune system appears to be modulated, not only by feedback mechanisms mediated through neural and endocrine processes, but by feedforward mechanisms as well. The immunologic effects of learning, an essential feedforward mechanism, suggest that, like direct neural and endocrine processes, behavior can, under appropriate circumstances, serve an immunoregulatory function in vivo.(ABSTRACT TRUNCATED AT 400 WORDS)