In this study, we exposed adult rats to chronic variable stress (CVS) and tested the hypothesis that previous early-life exposure to stress changes the manner in which older subjects respond to aversive conditions. To this end, we analyzed the cytogenic changes in the hippocampus and hippocampal-dependent spatial learning performance. The experiments were performed on 18-month-old male rats divided into four groups as follows: Control (old rats under standard laboratory conditions), Early-life stress (ELS; old rats who were exposed to environmental noise from postnatal days, PNDs 21–35), CVS + ELS (old rats exposed to a chronic stress protocol who were previously exposed to the early-life noise stress) and CVS (old rats who were exposed only to the chronic stress protocol). The Morris Water Maze (MWM) was employed to evaluate the spatial learning abilities of the rats at the end of the experiment. Immunohistochemistry against 5′Bromodeoxyuridine (BrdU) and glial fibrillar acidic protein (GFAP) was also conducted in the DG, CA1, CA2 and CA3 regions of the hippocampus. We confocally analyzed the cytogenic (BrdU-labeled cells) and astrogenic (BrdU + GFAP-labeled cells) changes produced by these conditions. Using this procedure, we found that stress diminished the total number of BrdU+ cells over the main proliferative area of the hippocampus (i.e., the dentate gyrus, DG) but increased the astrocyte phenotypes (GFAP + BrdU). The depleted BrdU+ cells were restored when the senile rats also experienced stress at the early stages of life. The MWM assessment demonstrated that stress also impairs the ability of the rats to learn the task. This impairment was not present when the stressful experience was preceded by the early-life exposure. Thus, our results support the idea that previous exposure to mild stressing agents may have beneficial effects on aged subjects.