1. The bark of the root and stem of various Magnolia species has been used in Traditional Chinese Medicine to treat a variety of disorders including anxiety and nervous disturbances. The biphenolic compounds honokiol (H) and magnolol (M), the main components of the Chinese medicinal plant Magnolia officinalis, interact with GABA(A) receptors in rat brain in vitro. We compared the effects of H and M on [3H]muscimol (MUS) and [3H]flunitrazepam (FNM) binding using EDTA/water dialyzed rat brain membranes in a buffer containing 150 mM NaCl plus 5 mM Tris-HCl, pH 7.5 as well as [35S]t-butylbicyclophosphorothionate (TBPS) in 200 mM KBr plus 5 mM Tris-HCl, pH 7.5. H and M had similar enhancing effects on [3H]MUS as well as on [3H]FNM binding to rat brain membrane preparations, but H was 2.5 to 5.2 times more potent than M. 2. [3H]FNM binding. GABA alone almost doubled [3H]FNM binding with EC50 = 450 nM and 200 nM using forebrain and cerebellar membranes, respectively. In the presence of 5 microM H or M the EC50 values for GABA were decreased to 79 and 89 nM, respectively, using forebrain, and 39 and 78 nM, using cerebellar membranes. H and M potently enhanced the potentiating effect of 200 nM GABA on [3H]FNM binding with EC50 values of 0.61 microM and 1.6 microM using forebrain membranes, with maximal enhancements of 33 and 47%, respectively. Using cerebellar membranes, the corresponding values were 0.25 and 1.1 microM, and 22 and 34%. 3. [3H]MUS binding. H and M increased [3H]MUS binding to whole forebrain membranes about 3-fold with EC50 values of 6.0 and 15 microM. Using cerebellar membranes, H and M increased [3H]MUS binding approximately 68% with EC50 values of 2.3 and 12 microM, respectively. Scatchard analysis revealed that the enhancements of [3H]MUS binding were due primarily to increases in the number of binding sites (Bmax values) with no effect on the high affinity binding constants (Kd values). The enhancing effect of H and M were not additive. 4. [35S]TBPS binding. H and M displaced [35S]TBPS binding from sites on whole rat forebrain membranes with IC50 values of 7.8 and 6.0 microM, respectively. Using cerebellar membranes, the corresponding IC50 values were 5.3 and 4.8 microM. These inhibitory effects were reversed by the potent GABA(A) receptor blocker R5135 (10 nM), suggesting that H and M allosterically increase the affinity of GABA(A) receptors for GABA and MUS by binding to sites in GABA(A) receptor complexes. 5. Two monophenols, the anesthetic propofol (2,6-diisopropylphenol, P) and the anti-inflammatory diflunisal (2',4'-difluoro-4-hydroxy-3-biphenyl carboxylic acid, D) also enhanced [3H]MUS binding, decreased the EC50 values for GABA in enhancing [3H]FNM binding and potentiated the enhancing effect of 200 nM GABA on [3H]FNM binding, although enhancements of [3H]MUS binding for these monophenols were smaller than those for H and M, using forebrain and cerebellar membranes. The enhancing effect of P and D on [3H]MUS binding were almost completely additive. 2,2'-biphenol was inactive on [3H]MUS and [3H]FNM binding. These, and other preliminary experiments, suggest that appropriate ortho (C2) and para (C4) substitution increases the GABA-potentiating activity of phenols. 6. The potentiation of GABAergic neurotransmission by H and M is probably involved in their previously reported anxiolytic and central depressant effects.