Achieving a functional cure is an important goal in the development of HIV therapy. Eliciting HIV-specific cellular immune responses has not been sufficient to achieve durable removal of HIV-infected cells due to the restriction on effective immune responses by mutation and establishment of latent reservoirs. Chimeric antigen receptor (CAR) T cells are an avenue to potentially develop more potent redirected cellular responses against infected T cells. We developed and tested a range of HIV- and SIV-specific chimeric antigen receptor (CAR) T cell reagents based on Env-binding proteins. In general, SHIV/SIV CAR T cells showed potent viral suppression in vitro, and adding additional CAR molecules in the same transduction resulted in more potent viral suppression than single CAR transduction. Importantly, the primary determinant of virus suppression potency by CAR was the accessibility to the Env epitope, and not the neutralization potency of the binding moiety. However, upon transduction of autologous T cells followed by infusion in vivo, none of these CAR T cells impacted either acquisition as a test of prevention, or viremia as a test of treatment. Our study illustrates limitations of the CAR T cells as possible antiviral therapeutics.