Two groups have tracked CD4+ T cells in infected mice to determine how memory CD4+ T cell populations arise in vivo. Pepper et al. identified two CD4+ T cell populations that were present in the early stages of Listeria monocytogenes infection and that gave rise to distinct memory subsets. The effector T cells of one population expressed T-bet and differentiated into T helper 1 (TH1)-type effector memory T cells in an interleukin-2-dependent manner. The other CD4+ T cell population expressed BCL-6 and CXCR5, depended on B cell-delivered signals through ICOS and gave rise to T central memory (TCM) cells. Thus, the precursor TCM cells closely resembled follicular helper T (TFH) cells, which promote B cell antibody responses. However, the TCM cells localized to the T cell areas of lymph nodes and did not retain BCL-6 expression. The authors suggest that CXCR5+ effector T cells that maintain BCL-6 expression may differentiate into TFH cells; the remainder may become TCM cells. Marshall et al. studied acute LCMV infection and also described two populations of effector CD4+ T cells with distinct memory potential. The CD4+ T cells of the first population were T-bethiLY6Chi and resembled terminally differentiated effectors. The second T cell population had a T-betmidLY6Cmidphenotype, was longer-lived and showed increased proliferation in response to secondary viral infection. This latter population showed a gene-expression profile that was remarkably similar to that of mature memory CD4+ T cells, suggesting that the T-betmidLY6CmidCD4+ T cells are memory cell precursors and gain memory cell attributes rapidly following virus infection.