This study investigates the possible involvement of opioid receptors in the action of a variety of anticonvulsant agents. The opioid antagonist naloxone (0.3, 1 mg/kg IP) and the selective mu-opioid antagonist cyprodime (3 mg/kg IP) significantly inhibited the increase in electroshock seizure threshold induced by phenytoin (3 mg/kg IP) in mice. The anticonvulsant effects of ethanol (1 g/kg IP) were also significantly antagonised by naloxone (1 mg/kg IP) but not by a 0.3 mg/kg IP dose or by cyprodime (3 mg/kg IP). The results with naloxone were confirmed using higher doses of phenytoin (10 mg/kg IP) and ethanol (1.5 g/kg IP). In contrast to the above findings, naloxone (0.3, 1 mg/kg IP) had no effect on the increase in seizure threshold induced by sodium valproate (200 mg/kg IP) or dizocilpine (MK801, 0.5 mg/kg IP) and paradoxically potentiated the increase in seizure threshold produced by phenobarbitone (15 mg/kg IP); carbamazepine (10 mg/kg IP) and the benzodiazepine agonist loprazolam (1 mg/kg IP), clearly differentiating these compounds from phenytoin and ethanol. These findings suggest that the anticonvulsant effects of phenytoin and ethanol (as assessed by their ability to prevent tonic hindlimb extension in the mouse electroshock model) may be mediated, at least in part, by the release of endogenous opioids and subsequent activation of opioid receptors (mu, in the case of phenytoin, but non-mu, in the case of ethanol) although direct activity at opioid receptors cannot be precluded.