Regulation of a memory IgE antibody response may be different from the induction of a primary response and may, therefore, be more relevant to the study of allergic diseases and the therapeutic manipula tion of IgE antibody formation. In this paper a murine hapten-specific in vitro memory IgE antibody response to benzylpenicilloyl (BPO)-KLH is described. The response was analyzed by determining the number of antibody-producing cells (APC) in an ELISA spot assay. Of the total number of BPO-specific APC (10,000 APC/106cultured spleen cells), about 1% were IgE-producing cells (100/106cultured cells), as detected on day 6 of culture. The level of the antibody response is antigen dose-dependent, and the detected APC are BPO specific. The memory IgE response is not inhibited by the addition of anti-IL-4 antibody (11B11), even at a high excess. In the presence of the mitogen lipopolysaccharide, it has been shown that switch of B cells to IgE is induced by IL-4, a process which can be inhibited by anti-IL-4 antibody. Because the antigen-induced IgE response cannot be inhibited by anti-IL-4 antibody, in vitro responding cells derived from BPO-KLH-preimmunized mice may, therefore, have already switched in vivo to IgE. On the other hand, B cells switching to IgE in a situ ation of cognate TB cell interaction might receive IL-4 in a transsynaptical way from T cells which might not be accessible to inhibition by anti-IL-4 antibody. The identification of the two possibilities in situations of esta blished allergic disorders will be decisive for determining whether pharmacological inhibition of IL-4 (or IL-4-induced switch)-eg, by putative low molecular weight compounds-will ever be a meaningful approach to suppress allergic diseases.