E arly blockade of T cell-costimulatory activation pathways prevents development of experimental chronic allograft rejection. Ongoing T cell recognition of alloantigen and activation may also play an important role in progression of chronic rejection, but definitive evidence is lacking. We used the fusion protein CTLA4Ig to block CD28-B7 T cell costimulation late after the onset of initial graft injury. Using the F334 into LEW rat model of chronic renal allograft rejection, transplant recipients were treated with a 10-d course of cyclosporine, and a subgroup received a single injection of CTLA4Ig at 8 wk after transplant. Functionally, CTLA4Ig administration prevented development of progressive proteinuria (14.3+/-4.1 mg/24 h versus 41.0+/-12.0 mg/24 h at 24 wk after transplant, P < 0.05). Histologically, graft mononuclear cell infiltration, glomerular hypertrophy, focal and segmental glomerulosclerosis, and intimal vascular hyperplasia were all attenuated in CTLA4Ig-treated animals. Lastly, reverse transcriptase-PCR and immunohistologic studies showed a significant reduction in the intragraft expression of key products of T cell and macrophage activation, and upregulation of what have recently been termed as "protective" genes, including the bcl family members, Bcl-2 and Bcl-xL, and hemoxygenase. Our data are the first to demonstrate that blocking T cell-costimulatory activation late after transplantation, after initial graft injury, prevents progression of chronic allograft rejection supporting the hypothesis that ongoing T cell recognition of alloantigen and activation are key mediators of ongoing chronic allograft rejection.