In recent studies we have demonstrated that immunological B- and T-cell memory may be stimulated effectively by oral immunization, simply by admixing protein antigens with cholera toxin (CT) adjuvant. Here we extend information by employing a hapten-carrier system allowing us to separate B- and T-cell memory and to evaluate the requirement of memory T cells for effective reactivation of mucosal memory B cells. We found that 2 weeks following oral priming immunizations with dinitrophenyl-keyhole limpet haemocyanin (DNP-KLH) plus CT adjuvant, significant serum anti-DNP antibodies of IgG, IgA and IgM immunoglobulin classes were demonstrated. However, after 2 years only IgM anti-DNP antibodies could still be detected in serum. When memory lymphocytes were isolated from these mice, from both systemic and gut-associated lymphoid tissues, and challenged with antigen in vitro, vigorous IgM, but no IgG or IgA, anti-DNP production was observed. By contrast, when the DNP-KHL-primed memory mice were challenged in vivo by an oral booster immunization with DNP-KLH plus CT adjuvant, strong systemic IgG and local mucosal IgA anti-DNP responses were recorded, while IgM anti-DNP production was poor. Moreover, the mucosal memory B cells from DNP-KHL-immunized mice were more responsive in vivo to an oral booster immunization with the carrier-specific antigen, DNP-KLH, compared to that provided by an unrelated carrier, DNP-human serum albumin (HSA), which gave only poor mucosal and systemic anti-DNP B-cell responses. Taken together our data suggest that mucosal memory B cells are recirculating cells that have retained their ability to produce IgM antibodies and, therefore, have not undergone switch differentiation involving gene rearrangements with constant mu-chain deletions. Furthermore, mucosal B-cell memory and CD4+ T-cell memory are closely interconnected phenomena, requiring both components for effective expression and probably also for maintenance of immunological memory in the mucosal immune system.