3alpha-OH-5alpha[beta]-pregnan-20-one (THP) is a positive modulator of the GABAA receptor (GABAR), which underlies its reported anxiolytic effect. However, there are conditions such as premenstrual dysphoric disorder (PMDD) where increases in THP levels can be associated with adverse mood. In order to test for conditions where THP might be anxiogenic, we developed a mouse model of THP withdrawal. Because delta-containing GABAR are highly sensitive to THP modulation, results were compared in wild-type and delta knockout mice. Finasteride, a 5alpha-reductase blocker, was administered for 3 days to female wild-type or delta knockout mice. Then, animals were tested in the elevated plus maze, following acute administration of THP, lorazepam, flumazenil, or 4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridin-3-ol (THIP), and results compared to vehicle-injected controls. CA1 hippocampal GABAR alpha4 subunit levels were assessed by Western blot. After THP withdrawal, THP produced anxiogenic effects, decreasing open arm entries on the elevated plus maze, following a brief shock, in contrast to its expected anxiolytic effects. As we have shown in rats, THP withdrawal also resulted in increased expression of the alpha4 subunit in mouse CA1 hippocampus. As expected for increases in alpha4-containing GABAR, THP withdrawn mice were relatively insensitive to the benzodiazepine (BDZ) lorazepam and had atypical responses to the BDZ antagonist flumazenil when tested on the plus maze. In contrast, they showed a greater anxiolytic response to THIP, which has greater efficacy at alpha4betadelta than other GABAR. Although THP withdrawal in delta knockout mice also increased the alpha4 GABAR subunit, the anxiogenic effects of THP and the anxiolytic effects of THIP were not observed, implicating alpha4betadelta GABAR in these effects. Based on these behavioral and pharmacological findings, we suggest that THP withdrawal in the mouse may serve as a rodent model of PMDD.