As the key molecule of type-I-hypersensitivity, IgE provides specificity for the allergen and links it to the allergic effector functions. Antibodies are secreted by plasma cells and their precursors, the plasma blasts. The fate of plasma cells is a subject of controversy, with respect to their lifetime and persistence in the absence of allergen. In general, plasma cells were for a long time considered as short-lived end products of B-cell differentiation, and many of them are short-lived, although already for more than 20 years evidence has been provided that IgE-secreting plasma cells can persist over months. Today long-lived, "memory" plasma cells are considered to represent a distinct cellular entity of immunological memory, with considerable therapeutic relevance. Long-lived plasma cells resist current therapeutic and experimental approaches such as immunosuppression, e.g. cyclophosphamide, steroids, X-ray irradiation, anti-CD20 antibodies and anti-inflammatory drugs, while the chronic generation of short-lived plasma cells is sensitive to conventional immunosuppression. The seasonal variation in pollen-specific IgE can be suppressed by immunotherapy, indicating that component of the IgE response, which is stimulated with pollen allergen is susceptible to suppression. Targeting of the remaining long-lived, allergen-specific plasma cells, providing the stable IgE-titers, represents a therapeutic challenge. Here we discuss recent evidence suggesting, why current protocols for the treatment of IgE-mediated allergies fail: Memory plasma cells generated by inhalation of the allergen become long-lived and are maintained preferentially in the bone marrow. They do not proliferate, and are refractory to conventional therapies. Current concepts target plasma cells for depletion, e.g. the proteasome inhibitor bortezomib, BAFF and APRIL antagonists and autologous hematopoietic stem cell transplantation.