Perioperative seizures have numerous potential etiologies. In general, when seizures occur during surgery, their onset often coincides with the introduction of a specific anesthetic or analgesic drug. Conversely, postoperative seizures are more commonly due to nonanesthetic causes. However, there have been reports of postoperative convulsions that appeared to be caused by anesthetic or analgesic drugs administered intraoperatively via inhalation or injection (e.g., intravenous, epidural, or peripheral nerve block). Some anesthetics appear to possess both proconvulsant and anticonvulsant properties. One possible factor is an inherent pharmacodynamic variability in the responsiveness of inhibitory and excitatory target tissues in the CNS. This is well illustrated by the anticonvulsant and proconvulsant effects of progressively higher doses of local anesthetic drugs. This variability in neuronal responsiveness could also explain the conflicting findings for low versus high doses of fentanyl and etomidate. Furthermore, biological variation in the individual patient's responsiveness to certain anesthetic drugs could be an additional contributory factor. Differing structure-activity relationships might also explain why some anesthetic agents possess both proconvulsant and anticonvulsant properties. Relatively minor modifications in a drug's structure can influence its affinity for a specific receptor site and its intrinsic pharmacologic activity. For example, when methohexital was first introduced, convulsions were commonly encountered in patients with and without a history of epilepsy. Subsequent fractionation of the original compound into its two isomeric forms resulted in the identification of the isomer primarily responsible for this convulsive activity. In its present formulation (Brevital; Eli Lilly, Indianapolis, Ind.), the epileptogenic properties of methohexital are limited to patients with psychomotor epilepsy. However, compared with thiopental, excitatory effects are still more common with methohexital. The excitatory effects of methohexital are presumably due to its methylated structure. The inhaled anesthetic flurothyl (hexaflurodiethyl) ether and the intravenous anesthetic ketamine also illustrate how subtle changes in stereoisomerism can result in significant changes in structure-activity relationships. Flurothyl, a fluorinated ether analogue, reliably produces convulsions in nonepileptic patients, whereas its structural isomer isoindoklon has not been associated with seizure activity. Other examples of isomer or structural analogue relationships that produce differential effects on neuronal hyperexcitability include enflurane-isoflurane and meperidine-normeperidine. In conclusion, the patient population (epileptic or nonepileptic), the method of documentation (EEG study or clinical observation), and the method of EEG analysis (cortical or depth electrodes) must be considered to properly analyze the proconvulsant and/or anticonvulsant properties of an anesthetic or analgesic drug.(ABSTRACT TRUNCATED AT 400 WORDS)