Evidence that an abnormality of excitatory neurotransmission may contribute to the epileptic phenomena in various animal and human syndromes is reviewed. Altered glutamate transport or metabolism may be a contributory factor in some genetic syndromes and enhanced responsiveness to activation of NMDA receptors may be significant in various acquired forms of epilepsy. Decreasing glutamatergic neurotransmission provides a rational therapeutic approach to epilepsy. Potent anticonvulsant effects are seen with the acute administration of NMDA antagonists in a wide range of animal models. Some competitive antagonists acting at the NMDA/glutamate site show prolonged anticonvulsant activity following oral administration at doses free of motor side effects and appear suitable for clinical trial.