Low (1 mM), but not 10 mM, concentrations of ethanol selectively potentiate current gated by α 4 β 2 δ subunit combinations of the gamma-aminobutyric acid type A (GABA A ) receptor, a subtype increased in hippocampus after withdrawal from progesterone in a rodent model of premenstrual anxiety. In the current study, we tested the hypothesis that the anxiolytic effect of ethanol would exhibit a similar dose-response effect by using the acoustic startle response (ASR) and elevated plus-maze as behavioral models. To this end, adult, female rats were tested (1) 24 h after removal of a progesterone-filled capsule implanted subcutaneously for 21 days (progesterone withdrawal) or (2) on the day of diestrus, a low hormone state. Low doses of ethanol (0.2-0.4 mg/kg) produced a significant 60%-70% decrease in the ASR only in animals undergoing progesterone withdrawal. However, higher doses of ethanol (0.8-1.2 g/kg) were ineffective in these animals, resulting in an "inverted U" ethanol dose effect similar to that observed at recombinant α 4 β 2 δ subunit combinations of the GABA A receptor. Consistent with these findings, significant 70% attenuation of the ASR was also achieved after progesterone withdrawal with 3 mg/kg of 4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridin-3-ol (THIP), a GABA A receptor partial agonist with greater potency at α 4 βδ receptors than at other known isoforms. In contrast, this partial agonist was not anxiolytic in control animals. These results support the suggestion that very low doses of ethanol are anxiolytic in a model of premenstrual anxiety, whereas higher, potentially sedative, doses are without effect. The results may be relevant for altered ethanol sensitivity during premenstrual syndrome, when increased ethanol consumption has been reported.