Both alpha-2-adrenergic agonists and benzodiazepines exert anxiolytic and sedative effects when administered as preoperative medications. Clinical effects achieved with a combination of drugs, representative of these classes of compounds, is greater than that which could be expected from a simple additive response. Therefore, we investigated the nature of the interaction between dexmedetomidine, the highly-selective alpha-2-adrenergic agonist, and midazolam in a series of in vivo and in vitro studies in rats. Rats were administered midazolam, dexmedetomidine, or a combination of midazolam and dexmedetomidine intravenously to derive three dose-response curves for loss of righting reflex (LRR). LRR was determined in rats in a rotating cage (4 rotations/min) by observing whether the rat failed to maintain its upright posture for greater-than-or-equal-to 15 s exactly 2.5 min after drug administration. The effect of either flumazenil (benzodiazepine receptor antagonist) or atipamezole (the alpha-2-adrenergic antagonist) on the LRR was also determined. A probit analysis was performed and an isobologram for the ED50 was derived to assess the nature of the interaction. Rat brain membranes were prepared for receptor binding assays using [H-3]-flumazenil and [H-3]-rauwolscine to characterize the benzodiazepine and alpha-2-adrenergic receptors, respectively. The ability of either midazolam or dexmedetomidine to displace the radiolabeled ligand from the alternative receptor was assessed. To detect a possible kinetic interaction between the two drugs, separate cohorts of rats were administered the two drugs individually or in combination at the combination ED50 doses. Blood was collected following decapitation exactly 2.5 min after completion of drug infusion, and concentrations of dexmedetomidine (radioreceptor assay) and midazolam (gas-liquid chromatography) were assayed. Midazolam and dexmedetomidine showed a significant synergism when administered together: midazolam ED50 was 1.3 % and dexmedetomidine ED50 16% of the ED50s needed if given separately. Although flumazenil attenuated the hypnotic response to midazolam, it was ineffective against dexmedetomidine-induced LRR. Conversely, atipamezole blocked the hypnotic response to dexmedetomidine but not to midazolam. In addition, we noted no "cross-displacement" by the agonists for the alternative receptor in the radiolabeled ligand binding studies. The dexmedetomidine and midazolam concentrations were similar whether given alone or in combination. These data strongly support a pharmacodynamic mechanism for the synergistic interaction between the benzodiazepine midazolam and the alpha-2-adrenergic agonist dexmedetomidine. The molecular components underlying this pharmacodynamic interaction do not include the drugs' receptor binding sites.