The transfer of human peripheral blood lymphocytes (huPBL) into mice with severe combined immune deficiency (SCID) has allowed for the generation of a model in which it is possible to examine human immune cell function in vivo. This model also provides the researcher an invaluable tool with which to examine human disease states and to evaluate potential therapies using human cells and reagents under more physiological conditions. However, it is clear that there are still significant obstacles that need to be addressed to optimize this model. The primary problem with the huPBL-SCID mouse is the tremendous variability in the extent of human cell engraftment which can vary from donor to donor and even from mouse to mouse using the same huPBL donor. This variability currently limits attempts to standardize this model among investigators and subsequently, various laboratories have employed different methods to optimize and ascertain the extent of human cell engraftment. In addition, there appears to be significant xenogeneic host-versus-graft and graft-versus-host reactions, as well as lethal EBV-induced B-cell lymphomas that can spontaneously arise which may further complicate the interpretation of human immunological studies performed in these mice. Despite these obstacles, the huPBL-SCID model has the potential to serve as a model for the examination of human lymphoid cells in an in-vivo setting and has been usefully applied to the study of various human disease states including AIDS, transplantations, autoimmunity and cancer.