Many studies have used repeated testing at different ages to study the age-related decline in learning and memory in animals. Generally, it has been assumed that performance is relatively independent of prior exposure(s). The present study tested the hypothesis that rats retain a significant memory of some aspects of spatial tasks for many months. Sprague-Dawley rats were trained in a repeated acquisition water maze task at 12 months of age, and retested in the same task at 24 months of age. Compared to 24-month-old naive rats, the pretrained rats initially performed significantly better on their first days of retesting at 24 months of age. In a second experiment, animals were trained for 2 weeks at 12 months of age and tested/retested for 2 weeks at the age of 24 months. Whereas the performance of the 24-month-old naive group improved over the 2 weeks of testing, the latencies of the pretrained animals were consistently shorter both initially, and throughout the testing. The performance of the pretrained rats in the second week of testing was similar to young rats. Together these data indicate that 24-month-old rats can learn spatial tasks well and that the performance of old rats in spatial tasks can be significantly influenced by prior exposure to that task, even 12 months earlier. © 2002 Elsevier Science B.V. All rights reserved.