The purpose of this study was to compare the effects of immune T cells derived from mucosal and systemic tissues on specific humoral responses to the periodontal pathogen Porphyromonas gingivalis and to determine the effect these responses have on periodontal bone loss in an experimental rat model. T cells were derived from either the spleen or Peyer's patches of conventional Fischer rats given live P. gingivalis by the oral route and were adoptively transferred into groups of nude Fischer rats. A group of nude rats to which no T cells were transferred served as control. Within 24 h of the adoptive cell transfer, animals from all groups were orally challenged with live P. gingivalis. Serum and salivary responses were seen in rats given immune T cells, whereas a low, variable serum IgM and no serum IgG or salivary IgA response was detected in the control nude rats, indicating that the major host responses to P. gingivalis whole-cell antigens were T-cell dependent. Rats given immune Peyer's patch-derived T cells exhibited a different serum IgG and salivary IgA immune response pattern than animals given immune splenic T cells. A higher serum IgG, especially IgG2b, antibody response and less horizontal bone loss was seen in rats given splenic than Peyer's patch-derived T cells. Although some differences were seen in the levels of salivary IgA early in the study in the two experimental groups, the levels of IgA were comparable by the termination of the experimental period. Furthermore, no difference was seen in the amount of vertical bone support in these two groups, which was greater than in the controls. These results suggest a role for serum IgG and salivary IgA in periodontal disease protection. The presence of high serum IgG2b antibody activity suggest the involvement of Th1-like cells in the response, whereas salivary IgA antibodies indicates the participation of Th2-1ike cells. These results suggest that the balance between Th1 and Th2-1ike cells within the systemic and mucosal compartments and the humoral immune responses to P. gingivalis they mediate are important in determining whether the responses induced are protective against experimental bone loss after oral challenge with the periodontal pathogen P. gingivalis.