The non-selective alpha-amino-3-hydroxy-5-methyl-isoxazole-4-propionic acid (AMPA) receptor antagonists, 2,3-benzodiazepine derivatives CFM-1 (3,5-dihydro-7,8-dimethoxy-1-phenyl-4H-2,3-benzodiazepin-4-one) and CFM-2 (1-(4'-aminophenyl)-3,5-dihydro-7,8-dimethoxy-4H-2,3-benzodiazepin -4-one), following intraperitoneal (i.p.) administration, were studied against audiogenic seizures in genetically epilepsy-prone rats (GEPRs) or pentylenetetrazole induced kindling in rats. After acute i.p. administration the ED50 values of CFM-1 against the clonic and tonic phases of the audiogenic seizures 30 min after pretreatment were 40 (16-100) and 13 (8-25) micromol kg(-1), respectively. The animals used for chronic study were treated i.p. daily (at 10 h) for 4 weeks with CFM-1 (20 or 50 micromol kg(-1)). Chronic treatment for 2 weeks with CFM-1 gave ED50 values against clonic and tonic seizures of 39 (22-69) and 16 (8-25) micromol kg(-1), respectively, whereas chronic treatment for 4 weeks gave ED50 values against clonic and tonic seizures of 42 (18-98) and 17 (7-41.3) micromol kg(-1), respectively. The duration of anticonvulsant activity observed between 0.5 and 4 h following administration of CFM-1 was similar for acute and chronic treatment. Two groups of Sprague-Dawley rats received CMF (20 or 50 micromol kg(-1)) 30 min before a subconvulsant dose of pentylentetrazole (25 mg kg(-1) i.p.) which is able to increase seizure severity in control animals (i.e., chemical kindling). Pretreatment with CFM-2 delayed the progression of seizure rank during repeated administration of pentylentetrazole. At the end of the period of repeated pentylentetrazole treatment (6 weeks) the mean seizure score was 0 in vehicle treated controls, 4.3 in animals treated with vehicle + pentylentetrazole, 2.2 in rats treated chronically with CFM-2 (20 micromol kg(-1) i.p.) + pentylentetrazole and 1.0 in rats treated repeatedly with CFM-2 (50 micromol kg(-1) i.p.) + pentylenetetrazole. CFM-2 was also able to antagonize the long-term increase in sensitivity of the convulsant effects of GABA function inhibitors in pentylentetrazole-kindled animals. Thus, the administration of a challenge dose of pentylentetrazole (15 mg kg(-1) i.p.) or picrotoxin (1.5 mg kg(-1) i.p.) 15 or 30 days after the end of the repeated treatment showed that animals treated with CFM-2 were significantly protected against seizures induced by pentylentetrazole or picrotoxin. The data suggest that, following repeated treatment, tolerance to the novel AMPA receptor antagonists does not develop (CFM-1 in genetically epilepsy-prone rats and CFM-2 in the pentylentetrazole kindling model of epilepsy). Thirteen minutes after drug injection on days 1, 14 and 28 of chronic treatment the motor impairment induced by these compounds was studied with a rotarod apparatus. The TD50 values for CFM-1 or CFM-2-induced impairment of locomotor performance were similar following acute and repeated treatment. The data also suggest that some novel 2,3-benzodiazepines may have clinical potential for some types of epilepsy.