Numerous findings suggest that the substantia nigra pars reticulata (SNR) acts as a seizure-gating mechanism and constitutes a site where protection against a broad spectrum of seizures can be obtained by inhibiting the activity of GABAergic output neurons to SNR target regions, particularly the superior colliculus. Most pharmacological studies on SNR neurons have been conducted with local microinjection of drugs, whereas studies with systemic administration of anticonvulsant drugs, e.g., the antiepileptic drug valproic acid (VPA), have yielded inconsistent data. In the present study, we examined the dose response and time course of the effects of anticonvulsant doses of VPA and its major active metabolite E-2-en-VPA on extracellularly recorded spontaneous single unit activity of nondopaminergic SNR neurons in a large group of chloral hydrate-anesthetized rats. The activity of each neuron was recorded over a period of at least 30 min after intravenous drug or vehicle administration. Both drugs rapidly decreased the firing rate of SNR neurons. Compared with predrug base-line values of neuronal firing of SNR neurons, the average peak inhibition of neuronal activity was approximately 16% at the lowest dose (50 mg/kg), approximately 20% (VPA) or 30% (E-2-en-VPA) at 100 mg/kg and approximately 50% (VPA) or 60% (E-2-en-VPA) at the highest dose (200 mg/kg) tested. ED50 values calculated from these data were 240 mg/kg for VPA and 159 mg/kg for E-2-en-VPA, respectively. Determination of drug levels in SN at time of peak effect after administration of 100 mg/kg VPA or E-2-en-VPA indicated that E-2-en-VPA exhibited its inhibitory effect on SNR firing at significantly lower concentration than VPA, suggesting that E-2-en-VPA is more potent than VPA in this regard. Except one SNR neuron, all neurons recorded in 52 rats reacted with a reduction in firing rate to administration of VPA or E-2-en-VPA, demonstrating a rather homogeneous susceptibility of SNR neurons to these drugs. The inhibition of SNR neurons by these drugs could be a result of increased GABA turnover, which has previously been reported to be induced in this region by systemic administration of VPA. We conclude that the ability of VPA and E-2-en-VPA to reduce firing of SNR neurons could be critically involved in their effectiveness in inhibiting seizures in various experimental models.