Recent studies indicate that the time required for virus-infected cells to become vulnerable for the activity of CTL is of significance for the capacity of CTL to control ongoing viral reproduction. To investigate whether this applies to the effectiveness of HIV-1-specific CTL, we measured virus production in cultures containing CD4(+) T cells inoculated with HIV at low multiplicity of infection, and CTL directed against an early protein, Rev, or a late protein, RT. The Rev-specific CTL prevented at least 2 log(10) more HIV-1 production, in 10 days, than similar numbers of RT-specific CTL. To study how CTL effectiveness depends on variations in the potency of effector functions and kinetics of HIV protein expression, we developed a mathematical model describing CTL-target cell interactions during successive infection cycles. The results show that substantially higher CTL-mediated target cell elimination rates are required to achieve control as there is less time for CTL to act before infected cells release progeny virions. Furthermore, in vitro experiments with HIV recombinant viruses showed that the RT-specific CTL were at least as effective as the Rev-specific CTL, but only if the RT epitope was expressed as part of the early protein Nef. Together these results indicate that CTL control ongoing HIV reproduction more effectively if they are able to recognize infected cells earlier during individual viral replication cycles. This provides rationale for immunization strategies that aim at inducing, boosting or skewing CTL responses to early regulatory proteins in AIDS vaccine development.