We sought to describe quantitatively the morphological and functional changes that occur in the dentate gyrus of kainate-treated rats, an experimental model of temporal lobe epilepsy. Adult rats were treated systemically with kainic acid, and, months later, after displaying spontaneous recurrent motor seizures, their dentate gyri were examined. Histological, immunocytochemical, and quantitative stereological techniques were used to estimate numbers of neurons per dentate gyrus of various classes and to estimate the extent of granule cell axon reorganization along the septotemporal axis of the hippocampus in control rats and epileptic kainate-treated rats. Compared with control rats, epileptic kainate-treated rats had fewer Nissl-stained hilar neurons and fewer somatostatin-immunoreactive neurons. There was a correlation between the extent of hilar neuron loss and the extent of somatostatin-immunoreactive neuron loss. However, functional inhibition in the dentate gyrus, assessed with paired-pulse responses to perforant-pathway stimulation, revealed enhanced, and not the expected reduced, inhibition in epileptic kainate-treated rats. Numbers of parvalbumin- and cholecystokinin-immunoreactive neurons were similar in control rats and in most kainate-treated rats. A minority (36%) of the epileptic kainate-treated rats had fewer parvalbumin- and cholecystokinin-immunoreactive neurons than control rats, and those few (8%) with extreme loss in these interneuron classes showed markedly hyperexcitable dentate gyrus field-potential responses to orthodromic stimulation. Compared with control rats, epileptic kainate-treated rats had larger proportions of their granule cell and molecular layers infiltrated with Timm stain. There was a correlation between the extent of abnormal Timm staining and the extent of hilar neuron loss. Granule cell axon reorganization and dentate gyrus neuron loss were more severe in temporal vs. septal hippocampus. These findings from the dentate gyrus of epileptic kainate-treated rats are strikingly similar to those reported for human temporal lobe epilepsy, and they suggest that neuron loss and axon reorganization in the temporal hippocampus may be important in epileptogenesis.