Unlike benzodiazepines, etifoxine appears to produce its anxiolytic effects by activating β2 and β3 subunit containing channels of the GABA A receptor complex (a different binding site than benzodiazepines), and by stimulating the production of GABA(A) active neurosteroids that act in conjunction with etifoxine's direct effects.  This difference in binding means that etifoxine can be used alongside benzodiazepines to potentiate their effects without competing for binding sites;  however, it also means that the effects of etifoxine are not reversed by the benzodiazepine antagonist flumazenil . 
In the mid 1980s, the neuroactive steroids 3α,5α-tetrahydroprogesterone or allopregnanolone (3α,5α-THP) and 3α,5α- tetrahydrodeoxycorticosterone (3α,5α-THDOC) were shown to modulate neuronal excitability via their interaction with GABA A receptors. The steroids 3α,5α-THP and 3α,5α-THDOC were able to enhance the GABA-elicited Cl − current.  In addition, these steroids might enhance the binding of muscimol and benzodiazepines to GABA A receptors.  Structure- activity studies (SAR) showed that the 3alpha-OH group is essential for the anesthetic actions of these steroids,  they also have an optimally-placed hydrogen bond accepting group on the β face of the steroid at the C-17 position. The four steroid rings form a rigid framework for positioning these hydrogen groups in three-dimensional space.  Analogues 5 and 6 (Figure 10) are weak modulators of GABA A receptor function because the flexible side chains in these analogues do not have the conformations required for high biological activity.