The discovery that T-cell recognition of antigen can have distinct outcomes has advanced understanding of peripheral T-cell tolerance, and opened up new possibilities in immunotherapy. Anergy is one such outcome, and results from partial T-cell activation. This can arise either due to subtle alteration of the antigen, leading to a lower-affinity cognate interaction, or due to a lack of adequate co-stimulation. The signalling defects in anergic T cells are partially defined, and suggest that T-cell receptor (TCR) proximal, as well as downstream defects negatively regulate the anergic T cell's ability to be activated. Most importantly, the use of TCR-transgenic mice was provided compelling evidence that anergy is an in vivo phenomenon, and not merely an in vitro artefact. The:je findings raise the question as to whether anergic T cells have any biological function. Studies in rodents and in man suggest that anergic T cells acquire regulatory properties; the regulatory effects of anergic T cells require cell to cell contact, and appear to be mediated by inhibition of antigen-presenting cell immunogenicity. Close similarities exist between anergic T cells, and the recently defined CD4(+)CD25(+) population of spontaneously arising regulatory cells that serve to inhibit autoimmunity in mice. Taken together, these findings suggest that a spec- tl um of regulatory T cells exists. At one end of the spectrum are cells, such as anergic and CD4(+)CD25(+) T cells, which regulate via cell-to-cell contact. At the other end of the spectrum are cells which secrete antiinflammatory cytokines such as interleukin 10 and transforming growth factor-beta. The challenge is to devise strategies that reliably induce T-cell anergy in vivo, as a means of inhibiting immunity to allo- and autoantigens.