CD28 and CTLA-4 are prototypal co-stimulatory and co-inhibitory cell surface signaling molecules interacting with CD80/86, known to be critical for immune response initiation and regulation, respectively. Initial “bench-to-beside” translation, two decades ago, resulted in the development of CTLA4-Ig, a biologic that targets CD80/86 and prevents T-cell costimulation. In spite of its proven effectiveness in inhibiting allo-immune responses, particularly in murine models, clinical experience in kidney transplantation with belatacept (high-affinity CTLA4-Ig molecule) reveals a high incidence of acute, cell-mediated rejection. Originally, the etiology of belatacept-resistant graft rejection was thought to be heterologous immunity, i.e., the cross-reactivity of the pool of memory T cells from pathogen-specific immune responses with alloantigens. Recently, the standard view that memory T cells arise from effector cells after clonal contraction has been challenged by a “developmental” model, in which less differentiated memory T cells generate effector cells. This review delineates how this shift in paradigm, given the differences in co-stimulatory and co-inhibitory signal depending on the maturation stage, could profoundly affect our understanding of the CD28/CD80-86/CTLA-4 blockade and highlights the potential advantages of selectively targeting CD28, instead of CD80/86, to control post-transplant immune responses.