Massive hepatic necrosis (MHN) is a condition that offers an opportunity to study the remarkable ability of the liver to become repopulated with hepatocytes. A maximal regenerative stimulus is expected in cases of MHN (Roskams et al. APMIS Suppl 1991;23:32-39). Sequential chronological observations, after a severe degree of liver cell loss, permit study of the human equivalent of the situation in animal models in which circulating and bone marrow-derived stem and liver progenitor cells are recruited to the hepatopoietic process. To date, the bone marrow and circulating precursors have not been identified morphologically in human material. We present data that suggest that the circulating liver progenitor could have a lymphoblastoid morphological appearance. Similar cells are seen among the cellular infiltrate of MHN. We have found that combinations of markers, such as CD117/CD133 positive CD45/tryptase negative are useful to isolate these cells using cell-sorting technology. This may facilitate their expansion in vitro and the development of their use for therapeutic purposes. In MHN, the residual portal tracts and ductular reaction with the associated lymphoid infiltrate (some of which are probably liver cell progenitors derived from the circulation) constitute the fundamental regenerative community unit in which hepatopoiesis takes place. Defining the hepatopoietic process is hindered by the lack of morphological transitional forms in the period between the progenitors within the circulation and when they assume recognizable hepatocytic form as "metaplastic" hepatocytes associated with the ductular reaction. By achieving a better comprehension of these processes of liver cell restoration, we will be better placed to accelerate liver recovery in MHN, for example by the administration of granulocyte colony stimulating factor (GCSF). Thus, more patients will be able to restore their own livers and avoid liver transplantation.