Current clinical gene therapy protocols for the treatment of human immunodeficiency virus type 1 (HIV-1) infection often involve the ex vivo transduction and expansion of CD4+ T cells derived from HIV-positive patients at a late stage in their disease (CD4 count <400). These protocols involve the transduction of T cells by murine leukemia virus (MLV)-based vectors encoding antiviral constructs such as the rev m10 dominant negative mutant or a ribozyme directed against the CAP site of HIV-1 RNA. We examined the efficiency and stability of transduction of CD4+ T cells derived from HIV-infected patients at different stages in the progression of their disease, from seroconversion to AIDS. CD4+ T cells from HIV-positive patients and uninfected donors were transduced with MLV-based vectors encoding beta-galactosidase and an intracellular antibody directed against gp120 (sFv 105) or Tat. (sFvtat1-Ckappa). The expression of marker genes and the effects of the antiviral constructs were monitored in vitro in unselected transduced CD4+ T cells. Efficiency and stability of transduction varied during the course of HIV infection; CD4+ T cells derived from asymptomatic patients were transducible at higher efficiencies and stabilities than CD4+ T cells from patients with acquired immunodeficiency syndrome (AIDS). Expression of the anti-tat intracellular antibody was more effective at stably inhibiting HIV-1 replication in transduced cells from HIV-infected individuals than was sFv 105. The results of this study have important implications for the development of a clinically relevant gene therapy for the treatment of HIV-1 infection.