RNA interference (RNAi) gene therapy against HIV-1 by stable expression of antiviral short hairpin RNAs (shRNAs) can potently inhibit viral replication in T cells. Recently, a mouse model with a human immune system (HIS) was developed that can be productively infected with HIV-1. In this in vivo model, in which Rag-2(-/-)gamma(c)(-/-) mice are engrafted with human CD34(+)CD38(-) hematopoietic precursor cells, we evaluated an anti-HIV RNAi gene therapy. Human hematopoietic stem cells were transduced with a lentiviral vector expressing an shRNA against the HIV-1 nef gene (shNef) or the control vector. We observed normal development of the different cell subsets of the immune system. However, although initial transduction efficiencies were similar for both vectors, a reduced percentage of transduced human immune cells was observed for the shNef vector after establishment of the HIS in vivo. Further studies are required to fully evaluate the safety implications. When we infected the mature human CD4(+) T cells from the HIS mouse ex vivo with HIV-1, potent inhibition of viral replication was scored in shNef-expressing cells, confirming efficacy. When challenged with an shNef-resistant HIV-1 variant, equal replication was scored in control and shNef-expressing cells, confirming sequence-specificity of the RNAi therapy. We thus demonstrated that an antiviral RNAi-based gene therapy on blood stem cells leads to HIV-1-resistant T cells in vivo, an important proof of concept in the clinical development of RNAi against HIV-1.