It has become widely accepted in the past decade that Th I and Th2 cells represent alternate states of function and gene expression of CD4 I T cells (MOSMANN and COFFMAN 1989; ABBAS et al. 1996). A question central to understanding the relevance of these subsets is whether they are products of an irreversible differentiation process or whether Th I and Th2 cytokine patterns can be interchanged in an ordered or a regulated manner. Many disease states can be attributed to the activity of one specific Th subset, such as Th I-mediated autoimmune diseases or Th2-mediated allergic diseases and this implies that the ability to alter or reverse Th differentiation is a potential strategy for the treatment of such diseases. The earliest evidence that Th I and Th2 cytokine patterns were stable in vitro came from the behavior of mouse Th clones in vitro (MOSMANN et al. 1986; MOSMANN and COFFMAN 1987). Many laboratories recognized that cytokine production patterns of most Th clones remained stable after repeated passage, and, in