The murine gamma-herpesvirus 68 (MHV-68) provides a unique experimental model for dissecting immunity to large DNA viruses that persist in B lymphocytes. The analysis is greatly facilitated by the availability of genetically disrupted (-/-) mice that lack key host-response elements, and by the fact that MHV-68 is a lytic virus that can readily be manipulated for mutational analysis. The mutant virus strategy is being used, for example, to characterize the part played in vivo by an MHV-68-encoded chemokine-binding protein that may ultimately find an application in human therapeutics. Experiments with various -/- mice and monoclonal antibody depletion protocols have shown very clearly that type I interferons (IFNs) are essential for the early control of MHV-68 replication, while CD4+ T cells producing IFN-gamma function to limit the consequences of viral persistence. Virus-specific CD8+ effectors acting in the absence of the CD4+ subset seem initially to control the lytic phase in the lung following respiratory challenge, but are then unable to prevent the reactivation of replicative infection in epithelia and the eventual death of CD4+ T-cell-deficient mice. This could reflect the fact that the interaction between the CD8+ T cells and the virus-infected targets is partially compromised by the MHV-68 K3 protein, which inhibits antigen presentation by MHC class I glycoproteins. Immunization strategies focusing on the CD8+ T-cell response to epitopes expressed during the lytic phase of MHV-68 infection can limit virus replication, but are unable to prevent the establishment of latency. Other experiments with mutant viruses also suggest that there is a disconnection between lytic MHV-68 infection and latency. The massive nonspecific immunoglobulin response and the dramatic expansion of Vbeta4+ CD8+ T cells, which is apparently MHC independent, could represent some sort of 'smoke screen' used by MHV-68 to subvert immunity. Although MHV-68 is neither Epstein-Barr virus nor human herpesvirus-8, the results generated from this system suggest possibilities that may usefully be addressed with these human pathogens. Perhaps the main lesson learned to date is that all the components of immunity are likely to be important for the control of these complex viruses.