The molecular bases for species barriers to lentiviral replication are not well understood, but are of interest for explaining lentiviral pathogenesis, devising therapeutic strategies, and adapting lentiviruses to gene therapy. HIV-1-based lentiviral vectors efficiently transduce nondividing cells, but present complex safety concerns. Nonprimate (ungulate or feline) lentiviruses might provide safer alternatives, but these viruses display highly restricted tropisms, and their potential for adaptation as replication-defective vectors capable of transducing human cells is unknown. Feline immunodeficiency virus (FIV) does not infect humans or other non-Felidae despite prevalent natural exposure. Although long terminal repeat (LTR)-directed FIV expression was found to be negligible in human cells, promoter substitution enabled an env-deleted, three-plasmid, human cell-FIV lentiviral vector system to express high levels of FIV proteins and FIV vectors in human cells, thus bypassing the hazards of feline vector producer cells. Pseudotyped FIV vectors efficiently transduced dividing, growth-arrested, and postmitotic human targets. The experiments delineate mechanisms involved in species-restricted replication of this lentivirus and show that human cells support both productive- and infective-phase mechanisms of the FIV life cycle needed for efficient lentiviral vector transduction. Nonprimate lentiviral vectors may offer safety advantages, and FIV vectors provide unique experimental opportunities.