Methodology for delivery of antigen to the lower respiratory tract of rhesus monkeys and the recovery of respiratory secretions (RS) by bronchial lavage has been evaluated and has been done on repeated occasions without apparent risk to animals. Rhesus immunoglobulin (Ig)G, IgM, and IgA were identified in RS by their cross-reactivity with human IgG, IgA, and IgM, and relative concentrations of rhesus IgA and IgG could be evaluated in sequential samples of RS and serum (S) by radial immunodiffusion analysis using antiserum specific for human gamma and alpha chains. Administration of keyhole limpet hemocyanin (KLH) by aerosol was compared with intravenous administration of KLH. The aerosol route of administration resulted in IgG, IgM, and IgA antibodies detectable by radio-immunodiffusion in RS but only IgG antibodies in serum. After separate sequences of exposure to KLH delivered to the lung, a secondary type of immune response in the lung appeared as indicated by time of appearance and height of antibody titers. Antibody titers against KLH were estimated by passive hemagglutination of tanned sheep red cells coated with KLH, and anti-KLH titers were comparatively higher in RS of animals immunized by the respiratory route. Treatment of RS and S with 2-mercaptoethanol resulted in reduction of hemagglutination titers of both S and RS, but the degree of reduction varied in different samples and was independent of whether or not the S and RS were collected simultaneously. These results indicate that at least part of the antibody response to aerosolized KLH in rhesus monkeys occurs in the lung and that antibodies may appear in all three immunoglobulin classes in this response.