The effect of two means of inducing a stress response, heat and oxygen radicals, on the ability of an HLA-DR1 B-cell line to stimulate DR1-restricted and anti-DR1 auto- and alloreactive T-cell clones has been examined. Both forms of stress enhanced the ability of B cells to stimulate auto- and alloreactive T-cell clones and to present peptide to an influenza-virus specific T-cell clone. Furthermore, the ability of the B-cell line to present whole influenza virus was augmented by heat stress. The stress-induced enhancement of T-cell responses coincided with a modest increase in the cell-surface expression of major histocompatibility class II products. This was, however, insufficient to account for the observed functional effects. In contrast to these effects, presentation of whole antigen was inhibited by the oxygen radical intermediate, hydrogen peroxide (peroxide), in a dose-dependent manner. When analysed by SDS-PAGE, it was found that whilst overall protein synthesis decreased following both types of stress, increased synthesis of heat-shock proteins (HSP), and in particular the 70,000 MW HSP, was only evident following heat stress. The absence of an increase in the synthesis of HSP 70, in the antigen-presenting cells (APC) following the uptake of UV-treated influenza virus, however, implied that HSP 70 induction was not necessary for the presentation of whole antigen. The effects of peroxide stress appeared to be qualitatively different in several respects. First, peroxide treatment did not cause the induction of any stress proteins; second, peroxide abolished the presentation of whole antigen. In addition, heat stress of APC was unable to protect from the adverse effects of peroxide treatment, in that cells treated sequentially with heat, followed by peroxide, were unable to present whole influenza virus. In order to determine the stage of antigen presentation at which peroxide was causing inhibition, APC were treated at varying time-points after pulsing with antigen. The kinetics of the peroxide effect paralleled those of aldehyde fixation. Taking these results together it would appear that peroxide interferes with some aspects of the antigen-processing pathway.