High doses of estradiol (E(2)) can impair spatial learning in the Morris water maze, in ovariectomized mice, but the same dose has no effect on adult castrated males. Here, we test the hypothesis that this sex difference is caused by neonatal actions of E(2). In Experiment 1, C57BL/6J pups were given daily estradiol benzoate (EB) or oil injections from the day of birth until postnatal Day 3. Adults were gonadectomized and received EB (s.c.) or oil 28 h before the first day of training, and 4 h before each of four daily training sessions on the Morris water maze. Females given oil as neonates, and EB prior to training displayed the poorest performance. Females that received EB as neonates and EB prior to training were insensitive to the deleterious effects of adult EB and performed better than males given the same hormone treatments. We conducted a second experiment using aromatase enzyme knockout (ArKO) mice. Adult male and female ArKO and wild-type (WT) littermates were gonadectomized and received either injections of oil or EB prior to and during water maze training (as described above). Hormone treatment failed to affect performance, yet, female but not male ArKO mice showed impaired learning compared to WT littermates. Thus, exposure to estradiol during neonatal development can counteract the deleterious effects of EB on adult spatial learning.