Social interactions as a consequence of the social position represent stressful conditions for the individual. Manipulation of social conditions or forming long-term social hierarchies by colony aggregation allow to investigate the regulation of immune defense mechanisms under seminatural circumstances. The present paper describes the effects of dyadic social interaction in male rats with or without direct aggressive interactions on some indices of humoral and cellular immunity. In addition, for comparative and reference purposes, in one experiment the conventional stressor of inescapable footshock was used as well. Primary humoral immune response to sheep red blood cell antigen is suppressed by repeated experience of both defeat and inescapable footshock. At individual level the social stressor is as effective as the conventional stressor of inescapable footshock, less individual rats show suppression following the social than after the conventional stress. The social stressors, i.e. being exposed to a resident or intruder of the territory, facilitate lymphocyte proliferation in the spleen by the mitogen ConA and PHA independently of the presence or absence of direct aggressive interaction. Finally, the different social stressors have some impact on the lymphocyte subpopulation in the spleen. Social stimulation without aggressive interactions increases the relative number of Thelper cells, whereas defeat leads to an increase in the Tsuppressor/cytotoxic subpopulation. The data suggest multiple and differential effects of social stress on immune system functioning in the rat. Individual characteristics of the coping with stress, the social environment, and the immune indices under investigation determine the magnitude and direction of the changes in immune functioning.