Langerhans cells (LCs) are bone marrow (BM) derived epidermal dendritic cells that represent a critical immunological barrier to the external environment, but little is known about their life cycle. We have recently discovered that in mice, LCs are maintained by local radio-resistant LC precursors that self-renew in quiescent skin throughout life but are replaced by circulating precursors upon inflammation (Merad et al. 2002). We have also shown that host radio-resistant LCs that persist after allogeneic HSCT play a major role in initiating skin GVHD (Merad et al. 2004). To explore the turnover of human LCs, we first asked whether human LCs are maintained by local LC precursors in quiescent skin of patients that have undergone HSCT. We assessed LC turnover from skin biopsies of recipients of sex-mismatched grafts and used X and Y chromosomes as a marker of host or donor origin. Patients skin samples were subjected to fluorescence in situ hybridization using X and Y specific DNA probes followed by immunofluorescence staining of Langerin (a LC specific marker) and nuclei staining. X and Y chromosomes were simultaneously visualized within the nuclei of Langerin-positive cells with high-resolution four-color microscopy. In a preliminary study, we found that in 7 patients with histological signs of skin GVHD, 98+3.4% of LCs were of donor origin. In contrast, in 2 patients that had no histological signs of GVHD, 100% LCs remained of host origin at 9 months and 4 years after myeloablative conditioning and HSCT despite 100% BM chimerism. These data suggest that in the absence of skin inflammation host LCs can persist in the skin for a long time despite complete donor hematopoietic engraftment. Next, we hypothesized that host LCs can persist in the skin because they are radio-resistant and able to proliferate in situ. Indeed, cell cycle analysis of purified LCs and co-staining of skin sections with Ki-67 and Langerin antibodies revealed that 2% of epidermal LCs present in quiescent human skin proliferate in situ, a percentage similar to that found in mice. To follow the behavior of human LCs experimentally, we reconstituted γ-irradiated Rag2 −/−γc−/− newborns mice with human CD34+ cord blood cells to engraft human hemato-lymphopoiesis. Engrafted animals developed functional human B cells, T cells and dendritic cells in the blood and spleen together with structured primary and secondary lymphoid organs. Consistent with our findings in patients, no human LCs were detected in the skin in steady state conditions, despite high levels of human hematopoietic cell engraftment. These results suggest that human LCs are maintained by a local pool of radio-resistant LC precursors that self-renew in situ. Since local LC precursors are eliminated during severe skin injuries, this xeno-transplantation model will be useful to dissect the homing molecules that govern the recruitment of LCs from the blood to the skin and may help develop new therapeutic strategies for the prevention and treatment of clinical GVHD.