Mice containing livers repopulated with human hepatocytes would provide excellent in vivo models for studies on human liver diseases and hepatotropic viruses, for which no permissive cell lines exist. Here, we report partial repopulation of the liver of immunodeficient urokinase-type plasminogen activator (uPA)/recombinant activation gene-2 (RAG-2) mice with normal human hepatocytes isolated from the adult liver. In the transplanted mice, the production of human albumin was demonstrated, indicating that human hepatocytes remained functional in the mouse liver for at least 2 months after transplantation. Inoculation of transplanted mice with human hepatitis B virus (HBV) led to the establishment of productive HBV infection. According to human-specific genomic DNA analysis and immunostaining of cryostat liver sections, human hepatocytes were estimated to constitute up to 15% of the uPA/RAG-2 mouse liver. This is proof that normal human hepatocytes can integrate into the mouse hepatic parenchyma, undergo multiple cell divisions, and remain permissive for a human hepatotropic virus in a xenogenic liver. This system will provide new opportunities for studies on etiology and therapy of viral and nonviral human liver diseases, as well as on hepatocyte biology and hepatocellular transplantation. Persistent infection with hepatitis B virus (HBV) is a major worldwide health problem, and chronically infected individuals are at high risk for developing cirrhosis and hepatocellular carcinoma. 1,2 Despite the availability of an HBV vaccine, there are still more than 350 million chronically infected people worldwide, and the few antiviral treatments currently available have a limited rate of efficacy. The narrow host range of HBV and the lack of both in vitro systems and of convenient animal models have greatly hampered our understanding of the complete virus life cycle, as well as the development of more effective antiviral drugs aimed at eradicating the virus from chronic carriers. 3 Chimpanzees are the only animal species infectable with HBV, 4,5 but studies with these animals and evaluation of antiviral therapies are severely restricted because of their limited availability and high costs. Animal models based on HBV-related hepadnaviruses, such as woodchuck and Pekin duck hepatitis B viruses, are often used for assessment of antiviral drugs 6-8 and have provided important information about factors involved in establishment of virus infection, viral persistence, and hepatocarcinogenesis. 9-14 However, woodchucks are relatively large animals of outbred origins that are difficult to handle in many laboratories, and chronic hepadnavirus infection in birds does not lead to cancer. The development of HBV-expressing transgenic mice has also provided important insights regarding viral pathobiology and the role of HBV gene products in hepatocellular injury. 12,[15][16][17][18][19] Although infectious virus can be produced in transgenic mice, their hepatocytes are not permissive for infection. Therefore, the still-unknown early step...