Administering estradiol to females during periods of low endogenous estradiol enhances their ability to extinguish fear, the laboratory basis of exposure therapy for anxiety disorders. It has therefore been proposed that estradiol could be a useful adjunct to enhance exposure therapy outcomes. The present study aimed to clarify the boundary conditions under which estradiol could be used for this purpose, by assessing whether the impact of estradiol, administered systemically prior to extinction training, differs depending on dose and estrous phase in adult female rats. Results demonstrated that in rats extinguished during metestrus (naturally low estradiol), a low dose of estradiol reduced freezing during extinction training and augmented extinction recall the following day, whereas a high dose of estradiol had no effect on either extinction training or recall. In rats extinguished during proestrus (naturally high estradiol), a high dose of estradiol impaired extinction recall, whereas a low dose of estradiol had no effect, or impairing effects, on extinction recall in different experiments. A subsequent analysis revealed that estradiol-treated proestrus rats that exhibited impaired extinction recall had significantly higher pre-treatment serum estradiol levels than those that exhibited good extinction recall. Together, these results indicate that systemically administered estradiol interacts with endogenous estradiol to produce an inverted U shaped dose effect on fear extinction, where low and high estradiol levels lead to poor extinction recall, and moderate estradiol levels lead to good extinction recall. These results highlight potential limitations to the use of estradiol as an adjunct to exposure therapy in clinical settings.