1. In chronically prepared, freely moving rats, electrical stimulation was applied to the perforant pathway and monosynaptic responses were recorded extracellularly in the ipsilateral dentate gyrus. In some tests a stimulus was also applied to the median raphe nucleus (mr) prior to activating the perforant pathway. Experiments were performed during two behavioral conditions: slow-wave sleep (SWS) and the still, alert state (SAL). Two varieties of evoked responses were recorded: those due to synchronous firing of neuronal action potentials (evoked action potentials or EAPs) and those produced by excitatory synaptic activity (evoked synaptic potentials or ESPs). 2. As reported previously (38), perforant path stimulation elicited EAPs of greater magnitude during SWS than during SAL. The application of a prior stimulus to mr (prestimulation) markedly increased the already elevated EAPs observed during SWS. The EAPs during SAL were unaffected by prestimulation. 3. The minimum delay time (time between mr and perforant path stimuli) at which the augmentation of the EAPs appeared during SWS was approximately 5 ms. The augmentation reached a maximum at delay times of 25-40 ms and was present up to a delay time of 150 ms. 4. As in former experiments (38), ESPs recorded in the molecular layer of the dentate gyrus after perforant path stimulation were found to be greater during SAL than during SWS. Prestimulation of mr had no significant effect on the ESPs at any level of the molecular layer during either SWS or SAL. 5. The perforant path afferent volley was recorded at high gain in the dentate gyrus. Its amplitude was found to be solely dependent on perforant path stimulus intensity and not on behavioral state or the prestimulation of mr. 6. In preparations anesthetized with Chloropent (82% chloral hydrate, 18% pentobarbital; Fort Dodge Laboratories, Fort Dodge, IA), prestimulation was applied at each of a number of loci within the pons and medulla, including mr, As in SWS, prestimulating mr resulted in augmented EAPs with a minimum delay time of 5 ms. Similar augmented responses were observed when stimulation was applied at other raphe nuclei (dorsal raphe, pontis, magnus, and pallidus), but there was no augmentation when stimulation was applied at other brain stem sites. Threshold stimulus intensities for producing augmented EAPs in the raphe nuclei were less than 30 microA. 7. In freely moving animals it was first established that the EAP responses during SWS were markedly greater than during SAL. Midline lesions were then made at the rostrocaudal level of mr. Following the lesions, there was no longer any significant difference in the magnitude of the EAPs recorded during the two behaviors. 8. These findings suggest that tonic influences arising from raphe nuclei during SWS may be involved in the facilitation of neuronal transmission through the dentate gyrus observed during this behavioral state.