Although the etiology of inflammatory bowel disease (IBD) remains unknown, a major working hypothesis is that it represents a dysregulated immune response to common enteric bacterial antigens. Until recently there has been a relative dearth of experimental models to study this hypothesis. However, exciting developments in experimental models of colitis, including spontaneous, transgenic and knockout mice, now allow this and other hypotheses to be tested. The regulation of mucosal immune responses is not well understood in the normal animal, much less in those with chronic intestinal inflammation. Clearly the CD4 Th1 and Th2 pathways are important in the host response to microbial pathogens, and recent data indicate that the intestinal mucosa seems to be a site of preferential Th2 responses towards exogenous antigens. Deletion of certain cytokine genes involved in maintaining this Th1/Th2 balance (interleukin [IL]-2, IL-10) resulted in colitis, although deletion of others (IL-4, interferon-gamma) that are also involved did not. Whether these cytokine gene deletions cause a dysregulation of the mucosal immune response has yet to be shown. However, the importance of regulation can be demonstrated in a model in which a normal CD4+ T cell subset (CD45Rb(high)) is transferred into syngeneic severe combined immunodeficiency syndrome recipients. This results in a striking colitis over the ensuing weeks with chronic diarrhea and wasting of the animals. If the reciprocal CD4+ subset (CD45Rb(low)) is co-transfered or if whole CD4+ T cells are transfered no colitis ensues. Therefore, T cells capable of causing colitis are present in normal animals but are prevented from doing so by immunoregulatory mechanisms. The antigens that drive the colitis in several of these models (IL-2 knockout mouse, human leukocyte antigen B27/β2M transgenic rat) appear to be those of the normal enteric bacterial flora because germ-free animals do not get the disease. Spontaneously colitic C3H/HeJBir mice also show prominent reactivity to enteric bacterial antigens. There are major differences among inbred mouse strains in susceptibility to colitis. The genes involved are not yet identified, but newly available technologies should allow that. In summary, these new models provide an experimental foundation to one of the major hypotheses on the cause of IBD, and will allow dissection of the genetic, environmental and immune components contributing to chronic colitis.