Immunity that is cross-protective between different influenza A virus subtypes (termed heterosubtypic immunity) can be demonstrated readily in some animals but only rarely in humans. Induction of heterosubtypic immunity in humans by vaccines would provide public health benefit, perhaps offering some protection against pandemics or other new influenza A strains. Therefore, we studied mechanisms mediating heterosubtypic immunity in mice. Immunization with either A/H1N1 or A/H3N2 virus protected mice against mortality following heterosubtypic challenge while providing modest reductions in lung virus titers. No cross-protection was seen with distantly related type B influenza virus. Depletion of CD4+ or CD8+ T cells or both around the time of challenge had no significant effect on survival, indicating that these cells are not required at the effector stage. beta2-microglobulin knockout mice could be protected readily against heterosubtypic challenge, confirming that class I-restricted T cells are not required. In beta2-microglobulin -/- mice, depletion of CD4+ T cells partially abrogated heterosubtypic immunity, showing that they play a role in these mice. Passive transfer of Abs to naive recipients protected against subsequent challenge with homologous but not heterosubtypic virus. Because a role for secretory Abs has been suggested, we studied dependence on the J chain, which is required for polymeric Ig receptor-mediated IgA transport. J chain knockout mice were readily protected by heterosubtypic immunity, indicating that polymeric Ig receptor-mediated transport is not required. Better understanding of heterosubtypic immunity should be valuable in analyzing new vaccines, including peptide and DNA vaccines, intended to induce broadly cross-reactive immunity.