We measured total IgG1, IgG2, IgG3, and IgG4 concentrations by ELISA in serum (S), total saliva (TS), cervicovaginal secretions (CVS), seminal secretions (SPE), and rectal secretions (RS) from either CDC II/III HIV-1-infected subjects or healthy volunteers. Human serum albumin was measured in parallel to calculate the relative coefficient of excretion (RCE). Levels of IgG1 and IgG3 directed against gp120 MN also were measured by ELISA in all samples, and the specific activity (SA) calculated. HIV-1-specific IgG2 and IgG4 were not compared, as total IgG2 and total IgG4 levels in HIV-1-infected subjects were found to be lower than in the healthy controls. Despite substantial interindividual variability, total IgG1 and IgG3 concentrations in all fluids were greater in the HIV-1-infected subjects than in the healthy controls. Calculations of RCE indicated predominantly a transudative origin for IgG subclasses in the different mucosal fluids, except for CVS, in which IgG1, IgG2, and IgG4 was produced locally. The transduction behavior of IgG3 in secretions appears to be different from that of other IgG subclasses. HIV-1-infected subjects were considered positive for IgG1 and IgG3 antibodies against gp120 MN if their antibody levels exceeded the maximum titer measured in the control group. Positive levels of anti-gp120 MN IgG1 were detected for 100% of HIV-1-infected individuals in S, CVS, and SPE, 97% in TS, and 75% in RS. Fewer subjects had positive levels of IgG3 to gp120 MN in their secretions (maximum 67% in CVS). Despite the low concentrations of total IgG3, mean SA values for IgG3 to gp120 MN were greater in secretions than in serum. No significant difference in the SA values for IgG1 to gp120 MN was observed between the different fluids. Only CVS had a local production of HIV-specific IgG1 Our results highlight the importance of an HIV-specific IgG1 and IgG3 immune response in mucosal fluids from HIV-1-infected subjects.