The relevance of EEG effect parameters as a measure of pharmacological effect intensity of benzodiazepines was evaluated. The concentration-EEG effect relationships of four benzodiazepine agonists, flunitrazepam, midazolam, oxazepam and clobazam, were quantified in individual rats and correlated with receptor affinity and anticonvulsant effect intensity of these compounds. Male Wistar-derived rats received a single i.v. dose of flunitrazepam (2.5 mg/kg), midazolam (5 mg/kg), oxazepam (10 mg/kg) and clobazam (20 mg/kg). Arterial blood samples were drawn frequently and EEG was monitored continuously until it had returned to preadministration levels. The concentrations of the benzodiazepines were determined by chromatographic means. Plasma protein binding was determined at 37 degrees C by ultrafiltration. The amplitudes in the 11.5 to 30 Hz frequency range, determined by aperiodic analysis, was used as EEG effect measure. Concentration-EEG effect relationships were derived by a pharmacokinetic-pharmacodynamic modeling procedure and characterized by the sigmoidal Emax model. The EC50 based on free drug concentrations (EC50,U, mean +/- S.E.) calculated for flunitrazepam (4.2 +/- 0.7 ng/ml) and midazolam (3.7 +/- 0.5 ng/ml) were similar and significantly less than the values for oxazepam (49 +/- 4 ng/ml) and clobazam (277 +/- 34 ng/ml) and illustrates the importance of using parameters referenced to unbound drug for comparative purposes. The maximal responses (Emax) for midazolam, oxazepam and clobazam were significantly less than for flunitrazepam suggesting that these three drugs may be regarded as partial agonists when compared to flunitrazepam. Receptor affinity was determined based on displacement of [3H] flumazenil in a washed brain homogenate at 37 degrees C.(ABSTRACT TRUNCATED AT 250 WORDS)