Corticosterone (CORT), the predominant glucocorticoid in rodents, elevated for 21 days damages hippocampal subregion CA3. We tested the hypothesis that CORT would impair spatial memory, a hippocampal function. In each of the three experiments, rats received daily, subcutaneous injections of either CORT (26.8 mg/kg body weight in sesame oil) or sesame oil vehicle alone (VEH). CORT given for 21 or 56 days effectively attenuated body weight gain and reduced selective organ and muscle weights. All behavioral testing was done on tasks that are minimally stressful and avoid deprivation. For each experiment, testing commenced 24h after the last injection. CORT given for 21 days did not impair spatial working memory in the Y-maze (Experiments 1 and 2). After 56-day administration of CORT, spatial working memory was impaired in the Y-maze (Experiment 2). CORT given for 21 days also failed to impair spatial working memory in the Barnes maze (Experiment 3). However, in trials that depended solely on reference memory, the VEH group improved in performance, whereas the CORT group did not. In conclusion, CORT elevated over a period of 21 days did not impair spatial working memory, but impaired the formation of a longer-term form of memory, most likely reference memory. Impairments in spatial working memory are seen only after longer durations of CORT administration.