In the present work we evaluated the anticonvulsant effects of two novel antagonists of the glycine co-agonist site (glycineB receptor) within the N-methyl-D-aspartate (NMDA) receptor complex, MRZ 2/576 (a tricyclic pyrido-phtalazin dione derivative) and L-701,324 (7-chloro-4-hydroxy-3-(3-phenoxy)phenyl-2(H)quinoline). As a model of epilepsy we used amygdala-kindled rats, which are considered as a model to study the efficacy of drugs against human complex partial seizures. MRZ 2/576 (2.5-10 mg/kg i.p. 15 min before testing) did not influence afterdischarge threshold, which is believed to be the most subtle indicator of efficacy against kindled seizures, nor did it affect other measures of seizure activity such as seizure severity, seizure duration and afterdischarge duration. However, MRZ 2/576 produced dose-dependent ataxia as measured in the open field and rotarod test. The highest dose tested (10 mg/kg) also markedly reduced rectal temperature (by about 1.5 degrees C). L-701,324 (2.5-10 mg/kg i.p. 30 min before testing) dose dependently and significantly increased afterdischarge threshold, but other seizure parameters remained unchanged. The ataxia produced by lower doses of L-701,324 (2.5 and 5 mg/kg) was more pronounced than that caused by MRZ 2/576. However, the ataxia observed following the higher dose of L-701,324 (10 mg/kg) was less intense than that elicited by MRZ 2/576. The behavioral alterations produced by the two drugs did not resemble those characteristic for classical competitive and uncompetitive NMDA receptor antagonists. In conclusion, our data indicate that glycineB receptor antagonists are not promising candidates for the treatment of complex partial seizures in humans, at least as monotherapy.