AbstractWe investigated the functional and behavioral implications of chronic corticosteroid removal in young and middle‐aged rats. Prepubertal and 13‐month‐old rats were adrenalectomized (ADX) or sham operated (SHAM). The young ADX rats were divided further into three groups: ADX with no hormone replacement, ADX given corticosterone chronically, (chCORT), and ADX given corticosterone acutely at the time of Morris water maze testing (acCORT). All rats were run on the Morris water maze 12 weeks after surgery. They were then sacrificed and the brains were removed for histological analysis.The results showed that prolonged corticosteroid absence caused major damage to the dentate gyrus and learning impairment on the Morris water maze. The chCORT rats had little dentate gyrus cell loss and were as efficient as the controls in Morris water maze performance, whereas the acCORT rats had dentate gyrus cell loss and were impaired in the spatial acquisition task. Furthermore, exogenously administered corticosterone had an interactive effect on ADX rats. Water maze performance was improved in dentate gyrus damaged rats (acCORT) compared to ADX rats not given corticosterone, whereas ADX rats with very little dentate gyrus damage (chCORT) did not exhibit better water maze performance relative to controls. Middle‐aged ADX rats lost cells only in the dorsal blade of the dentate gyrus but they did not show a learning impairment in the Morris water maze relative to the middle‐aged controls. These results indicate that corticosteroids are trophic for the dentate gyrus, that mature granule cells are less affected by adrenalectomy, that corticosteroid absence is responsible for some water maze impairment in ADX rats, but that in addition to corticosteroid absence, a substantial amount of dentate gyrus damage is necessary to impair spatial learning. © 1995 Wiley