When hippocampal slices are exposed to GABAA antagonists, area CA3 pyramidal cells and dentate hilar “mossy” cells discharge in synchronized epileptiform bursts (Muller and Misgeld, 1991; Scharfman, 1994b). Dentate interneurons are excited simultaneously, but the degree of discharge varies (Scharfman, 1994b). This study primarily examined the activity of dentate granule cells simultaneous to the epileptiform bursts of pyramidal cells and mossy cells. EPSPs followed by large GABAB receptor-mediated IPSPs were generated in granule cells during all epileptiform bursts of pyramidal cells and mossy cells, regardless of whether they were evoked or spontaneous. By simultaneous recording it was determined that granule cell EPSPs began several milliseconds after the start of pyramidal cell bursts (n = 48 simultaneous recordings) and immediately after the first action potential of a mossy cell burst (n = 77). Interneurons were similar to granule cells in the timing of their depolarizations relative to the onset of pyramidal cell (n = 24; Scharfman, 1994b) and mossy cell (n = 9) bursts. All excitatory activity was blocked by bath application of the glutamatergic AMPA/kainate receptor antagonist CNQX (5 microM, n = 5), but not the NMDA receptor antagonist D-APV (25–50 microM, n = 9). Granule cell EPSPs were decreased after focal application of CNQX to the molecular layer at a site close to the impaled granule cell (n = 5), whereas D-APV had no effect (n = 3). EPSPs also decreased after focal application of CNQX to the hilus, in two of four slices tested. The extracellularly recorded EPSP of granule cells was maximal in the inner molecular layer (n = 33), the site of the mossy cell axonal plexus. Severing the junction of the dentate gyrus and area CA3 blocked all spontaneous and evoked activity of dentate neurons without affecting burst discharges in area CA3a and CA3b (n = 6). None of the excitatory activity of any cell type was affected by cholinergic antagonists (atropine and mecamylamine, 25 microM each, n = 5; pirenzipine and dihydro-beta-erythroidine, 25 microM each, n = 5). The results suggest that there is a glutamatergic, AMPA/kainate receptor- mediated, excitatory pathway from area CA3 to the dentate gyrus in disinhibited slices. The pharmacological results, analyses of latency, as well as the known axonal projections of the sampled cells, suggest that the excitatory pathway begins within area CA3 and leads to granule cells via mossy cells. The data also suggest that dentate interneurons are excited by mossy cells, and possibly by pyramidal cells as well.(ABSTRACT TRUNCATED AT 400 WORDS)