We investigated the effects of two behavioral interventions--aerobic exercise and cognitive behavioral stress management (CBSM)--on Epstein-Barr virus viral capsid antigen (EBV-VCA) and human herpesvirus type-6 (HHV-6) antibody modulation in 65 asymptomatic gay men measured at several time points in the 5 weeks preceding and following notification of their human immunodeficiency virus-type 1 (HIV-1) serostatus. After accounting for potential immunomodulatory confounds, we found that HIV-1 seropositive men had higher EBV-VCA antibody titers than those diagnosed as seronegative at every time point during the study; however, no significant differences were found with respect to HHV-6. Among HIV-1 seropositive and seronegative subjects, respectively, those randomized to either behavioral intervention had significant decreases in both EBV-VCA and HHV-6 antibody titers over the course of the intervention as compared with assessment-only controls (of HIV-1 seropositive and seronegative status) whose antibody titers did not significantly change and which remained consistently higher than either serostatus-matched intervention group over subsequent time points, independent of total immunoglobulin G levels and degree of polyclonal B cell activation. In attempting to explain serostatus differences in EBV and HHV-6 values, it was found that HIV-1 seropositive men had significantly lower CD4 cells, CD4:CD8 ratio, and blastogenic response to phytohemagglutinin (PHA), as well as significantly higher CD8 cells at baseline. No significant differences were found between the HIV-1 seropositive and seronegative men with respect to anxiety and depression at baseline. Since the greatest changes in EBV and HHV-6 occurred between baseline and week 10, we correlated changes in immune (CD4, CD8, CD4:CD8 ratio, PHA stimulation) and distress-related markers (state depression and anxiety) with EBV and HHV-6 change scores over this time period. No significant correlations were found between any of these immune- or distress-related variable and the antibody change scores suggesting that the mechanisms by which EBV and HHV-6 antibodies are being modulated by these interventions possibly involve other, yet to be determined, immune, neuroendocrine, and/or psychologic variables.