Adult male rats were treated chronically with the selective type II corticosteroid receptor agonist dexamethasone, with dexamethasone plus aldosterone, a selective type I receptor agonist, and with a supraphysiological dose of corticosterone sufficient to occupy both type I and type II receptors; injection-free and oil (vehicle)-treated rats served as controls. Following one month of treatment, the animals were killed and their brains were processed for stereological assessment of volumes and total numbers of neurons in the hippocampal formation. Dexamethasone treatment resulted in significant reductions in the total number of dentate granule and the CA3 pyramidal cells and in the volumes of some layers of these subfields; however, this steroid did not influence any morphometric parameter in the CA1 subfield, and the number of hilar cells was also unaltered. In contrast to the results obtained with dexamethasone, the other two groups of corticoid injected animals did not reveal changes in total cell numbers in any of the subfields of the hippocampal formation, although in the corticosterone-treated group a reduction in the volumes of the hilus and of the stratum radiatum of the CA3 subfield was observed. The present data show that the exclusive activation of type II corticosteroid receptors results in subfield-specific neuronal loss in the hippocampal formation of rats. This type II receptor-mediated neuronal loss can, however, be abrogated by the simultaneous stimulation of type I corticosteroid receptors. Together, these findings extend and support previous studies which suggested that activation of type I corticosteroid receptors may promote neuronal survival and that neurodegeneration may be triggered by type II corticosteroid receptor stimulation. An important implication of this result is that elevated levels of the endogenous corticosteroid receptor ligands (e.g., during stress) is unlikely to cause severe structural damage to the hippocampal formation due to the contemporaneous occupation of type I receptors.