It has been 12 years since the first proposal was made to sub- divide mouse CD4 I T cell clones into Th I and Th2 subsets, based on their differences in cytokine production, and 7 years since the first clear demonstration of a similar dichotomy among human T cell clones. In the ensuing period, it has been realized that inappropriate development of Th I or Th2 responses are important features of many immunological and infectious dis- eases. Perhaps the first group of diseases to be understood in terms of preferential Th subset activation were allergic diseases (see PARRONC'HI et aI. , this volume). Several of the major, co- ordinately regulated, features of allergy, including IgE, eosino- philia and mastocytosis, were found to be stimulated by the T- specific cytokines I L-4 and IL-5 and inhibited by the Th I cyto- kine, IFN-. This suggested that the presence and severity of al- lergic responses reflected the relative numbers of Th I and Th2 cells specific for the offending allergen. Similarly, the very dif- ferent consequences of protective Th I and nonprotective Th2 responses to a number of intracellular pathogens have been re- cognized for some time (see TRINCHIERI and SCOTI, and COFF- MAN et aI. , this volume).