CHF3381 [n-(2-indanyl)-glycinamide hydrochloride] has been selected on the basis of a screening program as the compound displaying the highest anticonvulsant activity in the maximal electroshock seizure (MES) test and the best therapeutic index with reference to the rotarod test in mice and rats. In this study, the antiepileptic activity and the behavioural toxicity of CHF3381 were characterised in multiple model systems. CHF3381 effectively prevented MES-induced convulsions when administered i.p. (ED50, 24 mg/kg and 7.5 mg/kg) or p.o. (ED50, 21 mg/kg and 21 mg/kg) in both mice and rats, respectively. The time course of oral anti-MES activity in the rat was related to the brain concentration profile of unchanged CHF3381. Interestingly, the brain drug levels were about 4-5 times higher than in plasma. CHF3381 was very effective in mice against picrotoxin-, and i.c.v. N-methyl-D-aspartate (NMDA)-induced hind limb tonic extension (ED50 Approximately/=10 mg/kg), but was a weaker antagonist of 4-amynopyridine- and bicuculline-induced tonic seizures (ED50 approximately/=100 mg/kg), and ineffective against pentylentetrazole- and picrotoxin-induced clonic seizures. CHF3381 antagonised the behavioural effects and lethality of i.p. administered NMDA (ED50 = 57 mg/kg p.o.), indicating that the compound may act as a functional NMDA antagonist. In keeping with this idea, CHF3381 weakly displaced [(3)H]-TCP from binding to NMDA receptor channels (Ki, 8.8 microM). In the rat amygdala kindling model, CHF3381 was more efficient against kindling development than against kindled seizures (minimally active dose = 80 vs. 120 mg/kg i.p). Furthermore, it significantly increased the seizure threshold in kindled rats at relatively low doses (40 mg/kg i.p.). In contrast with MK-801-induced hyperactivity, CHF3381 moderately reduced the spontaneous locomotor activity in mice at anticonvulsant doses. Toxic effects on motor performance (rotarod test) were found at high doses only (TD50 approximately/= 300 mg/kg p.o., congruent with 100 mg/kg i.p. in both mice and rats). Furthermore, CHF3381 did not impair passive avoidance and Morris water maze responding in the therapeutic range of doses. Finally, the development of tolerance after repeated doses was negligible. These data indicate that CHF3381 exerts anticonvulsant and antiepileptogenic effects in various seizure models and possesses good therapeutic window, with scarce propensity to cause neurological side-effects.