In rats anesthetized with methoxyflurane, phenytoin (DPH) and medazepam (MDZ) were administered iontophoretically to pyramidal and granule cells discharging spontaneously or being driven by acetylcholine or glutamic acid. The objectives were to determine if these anticonvulsant agents exert direct effects on the rates of discharge of hippocampal neurons, whether similarities exist between responses elicited by DPH and MDZ; and if pyramidal and granule cells differ in their responsiveness to the drugs. The firing rates of 38% of spontaneously active neurons were reduced by iontophoretic DPH. The incidence of depression by DPH depended on the pretest discharge rates of the cells. Only 5% of cells with spontaneous rates less than 12/s were depressed by DPH, but 80% with rates faster than 12/s were inhibited. MDZ depressed 79% of spontaneously firing neurons regardless of pretest discharge rate. A majority of neurons whose firing rates were facilitated by either acetylcholine or glutamate were depressed by DPH or MDZ ejected iontophoretically. Pyramidal and granule cells responded similarly to putative transmitters, but differentially to the drugs. MDZ depressed a much greater proportion of spontaneously active granule cells than DPH. Phenytoin and MDZ differed with regard to the incidence of depression of spontaneous discharges, inhibition of slow firing cells, the proportion of granule cells depressed, and the duration of effect. These differences may be due to potency and pharmacokinetic factors or dissimilar mechanisms of action when the compounds are applied directly to single neurons.