The pharmacokinetic and pharmacodynamic interaction between the benzodiazepine agonist midazolam and antagonist flumazenil was quantified in vivo in rats, using effect parameters derived from aperiodic EEG analysis. The benzodiazepine-induced increase in amplitudes in the 11.5 to 30 Hz (beta) frequency band of the EEG was used as effect measure. A competitive interaction model was derived that could describe and predict the time course of EEG effect after administration of several combinations of midazolam and flumazenil. Two approaches to derive the interaction model were evaluated. In a first experiment, rats received 10 or 20 mg/kg midazolam i.v. in 15 min during a steady-state infusion of flumazenil at a rate of 0, 0.25, 0.5 or 1 mg/kg/hr. The EEG was continuously measured and frequent blood samples were taken to determine the pharmacokinetics of the drugs. Flumazenil did not influence the pharmacokinetics of midazolam, but a significant reduction in the clearance of flumazenil was observed in the presence of midazolam. Flumazenil did not possess any intrinsic efficacy with regard to the EEG effect measure. With increasing flumazenil concentrations, a parallel shift in the concentration-EEG effect relationship of midazolam was observed, confirming the competitive nature of the interaction. An agonist-antagonist interaction model was used to quantify the interaction, yielding the pharmacodynamic parameters of midazolam [(mean +/- S.E.): Emax = 80 +/- 5 microV/sec, EC50 = 35 +/- 3 ng/ml and N = 1.1 +/- 0.2] and flumazenil: EC50 = 24 +/- 2 ng/ml and N = 1.6 +/- 0.1. In a second experiment, rats received 5 mg/kg flumazenil or placebo i.v. in 10 min during a steadystate infusion of midazolam.(ABSTRACT TRUNCATED AT 250 WORDS)