Whether estrogen replacement is beneficial to cognitive health is controversial. Some studies have shown that estrogen replacement therapy (ERT) relieves memory impairment associated with menopause in women, whereas others suggest that estrogen not only is incapable of providing a benefit, but actually can be detrimental. One possible explanation for this discrepancy in study findings could be the varying time after menopause at which ERT is initiated. It has been proposed that a critical period exists during which ERT must be administered to enhance cognitive function. This idea has yet to be tested directly using functional synaptic studies, however. Here we investigated whether prolonged hormone deprivation caused by ovariectomy (OVX) in young adult rats prevents the ability of estrogen replacement to increase synaptic function in the hippocampus to a degree necessary for estrogen-induced improvement in learning and memory. Remarkably, estrogen replacement was found to increase long-term potentiation, the current mediated by NR2B-containing NMDA receptors, and the dendritic spine density at CA3-CA1 synapses up to 15 months post-OVX. However, by 19 months post-OVX, the same estrogen replacement was unable to induce these changes. Importantly, this loss of estrogen's effectiveness was seen to be a consequence of the duration of deprivation. In female rats aged with their ovaries intact and examined at the same chronological age as the 19-month post-OVX group, estrogen replacement significantly increased synaptic function and spine density. These data clearly demonstrate that a critical period exists during which ERT must be administered, and that once this period passes, the beneficial effects are lost.C ognitive function fluctuates across the menstrual cycle in women, with increased learning occurring when plasma estrogen levels are highest (1, 2). As such, loss of endogenous estrogen production after menopause, a consequence of normal aging, has been correlated with cognitive deficits (3). Treatment with estrogen replacement therapy (ERT) is not always successful in alleviating this hormone-related cognitive decline (4, 5). This lack of benefit of ERT may be explained by the critical period hypothesis, which proposes that there is a critical period after reproductive senescence during which estrogen is capable of increasing hippocampal function to a sufficient degree to enhance memory processing. After this period, ERT might be ineffective and possibly even detrimental (6, 7).Hippocampal learning is increased during proestrus in cycling rats, and this effect can be mimicked in young adult ovariectomized (OVX) rats treated with the ovarian estrogen 17β-estradiol (E2) (8-11). Similar to women, rats that experience long-term ovarian hormone deprivation no longer benefit from E2's effects on enhancing hippocampal learning (12-15). The role of ovarian hormone senescence in the effectiveness of E2 replacement in enhancing learning is difficult to distinguish from normal aging processes, given that aging alone is kn...