Despite its remarkable capacity for endogenous regeneration, the mammalian liver is vulnerable to a number of chronic or acute conditions that exceed or circumvent the proliferative capabilities of its mature cell complement. Bipotential hepatic progenitors, or "oval cells," have been shown to contribute to organ regeneration under such circumstances, both in human patients and in animal models. These progenitors are attractive agents for cell therapy, but have thus far proven challenging to isolate and manipulate. New reports indicating that transplanted bone marrow cells (BMCs) can also generate hepatocytes and contribute to liver repair have attracted considerable attention, because these cells are familiar and accessible to both clinicians and scientists. Recently, the issue of whether nuclear transfer (via cell fusion between donor BMC and recipient hepatocyte) or previously unrecognized differentiation potential (i.e., plasticity/transdifferentiation of BMC) is the primary origin of donor-derived hepatocytes has generated considerable controversy. In the liver, most evidence supports cell fusion as the key agent in the reversal of hepatopathology. However, regardless of their origin, the frequency of hepatocyte correction events is low. As is the case for the delivery of intrahepatic progenitors, substantial improvements in the understanding of this process will be needed before clinical application becomes practical.