One approach to gene therapy for hepatic diseases is to remove hepatocytes from an affected individual, genetically alter them in vitro, and reimplant them into a receptive locus. Although returning hepatocytes to the liver itself would be advantageous, the feasibility of this approach has never been evaluated due to the inability to distinguish donor from host hepatocytes. To unambiguously identify transplanted hepatocytes after transplantation, and to better quantitate their number and degree of liver function, two transgenic mouse lines were generated in a C57BL/6 background. The first expresses the Escherichia coli beta-galactosidase gene from the relatively liver-specific human alpha 1-antitrypsin (hAAT) promoter and allows transgenic hepatocytes to be readily identified after 5-bromo-4-chloro-3-indolyl beta-D-galactoside staining; the second produces the hAAT protein under control of the same promoter, which enables hepatocyte survival and maintenance of liver function to be quantitated by measuring the serum levels of hAAT. Hepatocytes isolated from transgenic donors were transplanted into nontransgenic C57BL/6 recipients by intrasplenic injection. Surprisingly, a large fraction of these cells were identified within the liver parenchyma but not the spleen at 2 months after transplantation. The high levels of serum hAAT detected in transplant recipients were stable for greater than 6 months, suggesting that established cells will survive indefinitely. These results have important implications for liver organogenesis and hepatic gene therapy.