Even though the effects of overtraining and glucocorticoids on different phases of spatial memory are known, the interaction between these factors on the retrieval and extinction of spatial memory has not yet been described. Adult male Wistar rats received eight training trials per day in the Barnes maze for either one or two days. Twenty-four hours after the last training trial they were randomly assigned for receiving an intraperitoneal vehicle or corticosterone injection (0.125 or 0.5 mg/kg) and ten minutes later they were given a memory test, followed by seven extinction trials. Extinction retention was evaluated twenty-four hours after extinction. The second training session did not provoke significant changes regarding escape latency nor weighted errors, thereby showing that overtraining had been obtained. The overtrained animals performed better than the trained ones during the retrieval test. Corticosterone administration did not affect the overtrained animals' performance; by contrast, only the lower dose impaired trained animals' retrieval. Overtrained subjects acquired extinction more rapidly than those which received just one session, but corticosterone did not significantly modify extinction. However, whilst the spatial task remained extinguished in trained animals during the extinction retrieval test, spontaneous recovery occurred in overtrained animals. Such training intensity effects on extinction retrieval were reverted by corticosterone. Overall, these results suggested that overtraining modified the susceptibility of spatial memory's trace to the effects of corticosterone on retrieval and extinction.