The incidence of asthma is at near-epidemic levels among the peoples of industrialized countries ( see review by Cookson and colleagues [1]). Hypotheses for this proliferation have included increases in air pollution, early exposure to trigger antigens (e.g., dust mites, cockroaches), genetic factors, and differences in the amount of time spent inside versus outside. Respiratory infections early in life (e.g., respiratory syncytial virus [RSV], rhinovirus, and parainfluenza) have also been proposed as predisposing factors toward the later development of asthma ( see Reference 2). In contrast, infections with organisms that induce a predominant T helper (Th)1 response (e.g., tuberculosis) have been hypothesized to provide a protective effect with regard to developing the Th2-driven syndromes of asthma and allergies (3). Despite intense study of patient populations, the unique circumstances that dictate why one person’s immune responses lead to asthma when others do not are still obscure. Since 1796, when Jenner showed that inoculation with cow pox conferred protection on the recipients, theories have been proffered to explain the immune response when an organism is confronted with an antigen. At the cellular level, distinct and alterable immune responses are attributed to a number of factors, including variations in the types of responding cells, the scope and intensity of the antigen encounter, and differences in the manner of antigen presentation ( see References 4 and 5). Furthermore, the cointroduction of antigen with an adjuvant is critical in determining the immune response obtained. In animal models of respiratory disease, a redundant and confounding factor has been the variability of responses obtained by even slight modifications of antigen-exposure protocols. Recent data from mouse models of allergic respiratory inflammation, however, implicate the history of antigen exposures as another critical variable in determining the type and magnitude of an immune response. A defining example is provided by the experiments of Erb and colleagues (6) who demonstrated that prior administration of the bovine Mycobacterium tuberculosis vaccine (bacillus Calmette-Guerin [BCG]) leads to the attenuation of the Th2 reactions generally associated with a subsequent ovalbumin (OVA) intranasal challenge. In contrast, animals with no previous exposure to BCG develop, as expected, a dramatic Th2 response to an OVA challenge. A prominent hypothesis to explain these results is that the prior immune history of an organism can change the cytokine milieu of a particular tissue or organ (in this case the lung), or that subpopulations of immune cells, such as T cells, have undergone a biased differentiation in view of prior antigen exposures and cytokine production (7, 8). In particular, the expression of interleukin (IL)-4 and interferon (IFN)g has been especially scrutinized because of these cytokines’ ability to promote the differentiation of Th2 cells or dramatically suppress Th2 responses, respectively (Table 1). At first glance, the proposal that the history of antigen exposures affects ongoing responses may seem like a logical extension of early theories of acquired immunity and immune memory. However, these previously mentioned studies extend the earlier theories by suggesting that immunologic memory is not confined to exposure to identical antigens. The significant extrapolation of this concept is that the immune response to the administration of a new “second” antigen may be dependent on previous exposure to any other antigens that have been encountered in the past. The regulation of these history-dependent responses at a cellular and/or molecular level could be accomplished in several ways. For example, antigen exposure leads to cellular responses that result in the generation of memory-effector T cells localized within particular tissues or lymphoid organs (Figure 1). The presence of these memory-effector cells in discrete lymphoid tissues has been recognized for some time, including the importance of their distinct functional capabilities and capacity to synthesize certain cytokines (9–11). The novelty suggested by recent reports is that the presence of particular cytokines in secondary lymphoid tissues alone may influence the type of response generated by irrelevant—and heretofore unseen—antigens. One explanation for this phenomenon at the cellular/molecular level is that a prior exposure to antigen causes specific cell populations to secrete particular cytokines that drive either Th1 (IL-12, IL-18, IFNg ) or Th2 responses (IL-4, IL-5, IL13) ( see Table 1). The presence of these polarizing cytokines may influence responses to subsequent antigen provocations, even if the subsequent antigens are not related to the first. The sources of these pulmonary cytokines may include a number of cell types in addition to memory-effector ( Received in original form July 9, 1999 )