Human-specific HIV-1 and hepatitis co-infections significantly affect patient management and call for new therapeutic options. Small xenotransplantation models with human hepatocytes and hematolymphoid tissue should facilitate antiviral/antiretroviral drug trials. However, experience with mouse strains tested for dual reconstitution is limited, with technical difficulties such as risky manipulations with newborns and high mortality rates due to metabolic abnormalities. The best animal strains for hepatocyte transplantation are not optimal for human hematopoietic stem cell (HSC) engraftment, and vice versa. We evaluated a new strain of highly immunodeficient nonobese diabetic/Shi-scid (severe combined immunodeficiency)/IL-2Rγc(null) (NOG) mice that carry two copies of the mouse albumin promoter-driven urokinase-type plasminogen activator transgene for dual reconstitution with human liver and immune cells. Three approaches for dual reconstitution were evaluated: i) freshly isolated fetal hepatoblasts were injected intrasplenically, followed by transplantation of cryopreserved HSCs obtained from the same tissue samples 1 month later after treosulfan conditioning; ii) treosulfan conditioning is followed by intrasplenic simultaneous transplantation of fetal hepatoblasts and HSCs; and iii) transplantation of mature hepatocytes is followed by mismatched HSCs. The long-term dual reconstitution was achieved on urokinase-type plasminogen activator-NOG mice with mature hepatocytes (not fetal hepatoblasts) and HSCs. Even major histocompatibility complex mismatched transplantation was sustained without any evidence of hepatocyte rejection by the human immune system.