A controlled experimental system for the evaluation of pulmonary immune responses in horses with "heaves" (recurrent airway obstruction) has been developed. We hypothesized that the humoral immune response to an inhaled antigen in diseased horses would be different from that of healthy horses and that chronic pulmonary inflammation would bias the production of IgG isotypes in diseased horses as compared to healthy horses. Healthy and affected horses were housed in a natural challenge environment (stabled, fed dusty hay) and exposed by inhalation, to a nebulized solution of keyhole limpet hemocyanin (KLH). Sera and bronchoalveolar lavage fluids (BALFs) were collected from horses prior to and following their inhalation exposure to the antigen. Differential cell counts were performed on the cells in the BALF. An enzyme-linked immunosorbent assay (ELISA) was used to determine the concentrations of IgGa, IgGb, IgG(T) and combined IgG specific for KLH in the sera and BALF. The percentages of neutrophils in the BALF of diseased horses were increased 4-6-fold over healthy horses. Combined IgG specific for KLH was significantly greater in BALF and serum from healthy compared to diseased horses. Differences in isotypes were also evident; however, only IgGb specific for KLH in the BALF was significantly increased in healthy versus diseased horses. Possible explanations for this difference include: (1) increased destruction of antigen before it could interact with lymphocytes, (2) down-regulation of IgGb production by inhibitory cytokines in diseased horses, or (3) binding of IgGb to Fc receptors on the large numbers of neutrophils in the lungs of diseased horses. In contrast to the prevailing notion that horses with heaves have exaggerated immune responses, our data suggest that diseased horses exposed to an aerosolized protein mount weaker IgG responses compared to healthy horses.