Rotavirus-naive and rotavirus-immune gnotobiotic calves were treated with high doses of dexamethasone (DX) to suppress the immune system. Calves were then infected with a virulent rotavirus inoculum, J-160, to investigate the role of immune responses both in recovery from primary rotavirus infection and in resistance to secondary rotavirus infection. Treatment of calves with DX markedly suppressed in vitro responsiveness of peripheral blood lymphocytes to mitogens within 48 h of the start of DX treatment. Suppression was similar in rotavirus-naive and rotavirus-immune calves. In contrast, the effect of DX treatment on specific antibody responses differed depending on when DX treatment started in relation to rotavirus infection. When DX treatment commenced prior to primary rotavirus infection both systemic and local specific antibody responses were inhibited. These calves, in which mitogen and antibody responses were suppressed, exhibited greater clinical signs than did control calves after infection with virulent rotavirus, but virus excretion was affected in only one of the two calves. When DX treatment was started after primary rotavirus infection but before secondary infection, systemic and local antibody responses to the primary infection and to the challenge infection were not affected. These calves resisted challenge with virulent virus as did DX-untreated rotavirus-immune calves, even though mitogen responses were suppressed. We conclude that in a primary rotavirus infection, virus excretion ceased when both antibody and mitogen responses were suppressed. Resistance to secondary rotavirus infection occurred when mitogen responsiveness was suppressed, but when antibody levels were normal. Thus, no evidence was obtained that fully functional cell-mediated immune mechanisms are essential for resistance to rotavirus infection. Evidence was provided for the ability of parenteral treatment with DX to suppress mucosal as well as systemic antibody responses.