Inflammatory bowel diseases take chronic courses due to the expansion of colitogenic CD4(+) cells. However, it is unclear whether the persistent disease is driven by continuous reactivation of colitogenic memory CD4(+) cells to generate effector CD4(+) cells or by continuous generation of effector CD4(+) cells from naïve cells. To clarify this issue, we performed a series of sequential adoptive transfers of Ly5.2(+) and Ly5.1(+) CD4(+)CD45RB(high) cells into RAG-2(-/-) mice at different time points. We show here that the secondarily transferred CD4(+)CD45RB(high) cells can be converted to CD4(+)CD44(high)CD62L(-)IL-7Ralpha(high) effector-memory T cells even in the presence of pre-existing effector-memory CD4(+) cells. Although the total cell numbers of CD4(+) cells in established colitic mice were consistently equivalent irrespective of the number of primarily transferred cells, the ratio of primarily and secondarily transferred cells was dependent on the ratio of the transferred cell numbers, but not on the order of the transfer. Of note, we found that primarily transferred CD4(+) cells produced significantly lower amounts of IFN-gamma and IL-17 than CD4(+) cells arising from secondary transfer. In conclusion, the continuous generation of colitogenic CD4(+) cells that compensate for exhausted CD4(+) cells may be one of the mechanisms involved in the persistence of colitis.