Estradiol benzoate (EB) has repeatedly been shown to increase hippocampal CA1 spine synapse density in ovariectomized female rats. Although this increase has been assumed to enhance memory, a direct link between increased spine synapse density and memory has not been demonstrated. Furthermore, while androgens, such as testosterone propionate (TP) and dihydrotestosterone (DHT), also increase spine synapse density in females, their effects on memory have yet to be investigated. In the present study, ovariectomized female rats were given two injections, 24 h apart, of sesame oil (control), 10 microg EB, 500 microg TP or 500 microg DHT. Forty-eight hours after the second injection, rats were tested in a 1-day spatial Morris water maze task and then immediately perfused for analysis of CA1 spine synapse density (using electron microscopy and unbiased stereology). In the spatial acquisition phase of testing, EB, but not TP or DHT, significantly impaired memory relative to controls. Hormone treatment did not affect spatial retention or performance in the non-spatial phase of testing. In contrast to previous work, spine synapse density was not increased by EB, TP or DHT. We therefore examined a new set of EB-treated females, only half of which were water maze tested. Consistent with previous work, EB significantly increased spine synapse density among behaviorally naïve females. In contrast, spine synapse densities did not differ among behaviorally tested control and EB females, although they were higher than behaviorally naïve controls. These data indicate that 1-day water maze testing can eliminate the hormone-induced increases in CA1 spine synapse density typically observed in behaviorally naïve females.