The past five or six years has seen a resurgence of interest in immunological memory. Areas in which important advances have been made of late or in which problems in understanding persist are covered here: (i) Selection of virgin B cells for entry into the peripheral pool. (ii) Expression of immunoglobulin isotypes and other markers on memory B cells. (iii) Development of memory B cells as a separate lineage from primary response B cells. (iv) Sites of production of memory B cells. (v) Signals that rescue mutating B cells in germinal centers, forming the basis of affinity selection, and programming further differentiation. (vi) The myriad markers of memory T cells, in particular CD45R isoforms. (vii) Selective migration pathways of memory T cells and its possible molecular basis. (viii) The lifespan of memory cells and factors that influence their long-term survival. The data accumulated during this period which have vastly increased our understanding of memory have at the same time highlighted unresolved problems that could block further progress in the field. The thorny question that we cannot at present answer is: How does a memory cell differ from an activated cell and, in the case of T cells, from an effector cell? The problem bears on the interpretation of any study that sets out to correlate memory phenotype with memory function. Immunologists may have donned an intellectual straitjacket in their search for the memory cell.