The importance of CD4+ CD25+ regulatory T cells (Treg) in maintaining immune homeostasis has been directly demonstrated in vivo by their manipulation in a number of autoimmune disease models in the mouse. In the study of human regulatory cells, we have found that the cells that consistently demonstrate the in vitro regulatory activity most similar to that described for murine cells in vitro are best identified by restricting the isolation of CD25+ CD4 T cells to those cells expressing only the highest levels of CD25, representing approximately 2-3% of total CD4 T cells. Thus, it is the CD4+ CD25high subset that exhibits the in vitro characteristics that are identical to the CD4+ CD25+ regulatory cells initially characterized in mice. Furthermore, the cells expressing medium to low levels of CD25 not only do not exhibit suppressive activity directly ex vivo, but also actually contain a significant proportion of CD62L- CD4 T cells which are believed to be in vivo activated T cells. Due to the inherent difficulties in using CD25 as a marker for the purification of Treg cells, the finding that selection of the CD25high subset of CD4+ CD25+ T cells minimizes the co-isolation of contaminating activated CD4 T cells is important for future studies of these Treg cells in human disease. In order to perform these studies, we first had to establish a highly reproducible 'micro in vitro co-culture' assay system to enable the functional analysis of high-purity, but low-yield regulatory populations derived from FACS sorting. With this system in place, we are poised to dissect the potential heterogeneity of mechanisms employed by highly specific subpopulations of CD4+ CD25+ cells.