The effectiveness of protection conferred by CD8(+) memory T cells is determined by both their quality and their quantity, which suggests that vaccine efficacy might be improved if it were possible to increase the size of the memory pool. Approximately 90% of virus-specific CD8(+) T cells die during the contraction phase and, herein, we have attempted to increase the memory pool by reducing CD8(+) T cell death. CD8(+) T cell contraction has been attributed to apoptosis, or programmed cell death (PCD), which, classically, is dependent on caspases. Caspase-dependent PCD can be prevented by the pan-caspase inhibitor benzyloxycarbonyl-Val-Ala-Asp (OMe)-fluoromethylketone (zVAD), and here we evaluate the effect of this compound on virus-specific T cell responses in mice. zVAD prevented caspase-dependent PCD of freshly isolated virus-specific T cells in tissue culture, and a fluorescent analog, FITC-VAD, entered CD8(+) T cells following in vivo injection. However, despite using 11 different regimens of zVAD administration in vivo, no significant effects on CD8(+) or CD4(+) memory T cell numbers were observed. Furthermore, the CD8(+) memory T cell responses to secondary virus infection were indistinguishable, both qualitatively and quantitatively, in zVAD-treated and normal mice. The absence of effect cannot be attributed to a technical flaw, because identical doses of zVAD were able to rescue mice from hepatocyte apoptosis and lethal intrahepatic hemorrhage, induced by inoculation of anti-Fas Ab. We conclude that the contraction phase of the virus-specific T cell response is unlikely to require caspase-dependent PCD. We propose that contraction can be mediated by an alternative, caspase-independent pathway(s)