Two experiments examined retention of spatial learning in rats using a Morris water maze. Retention was scored in terms of probe trial performance when the platform was removed. Latency to reach the platform location, percent of time in the quadrant that had contained the platform, and relative frequency of visits to the platform location were analyzed. Results of the first experiment showed that preweanlings and juveniles exhibited substantial forgetting at 3- and 7-day retention intervals. Forgetting in adults was much lower than that found in the younger animals, and no differences in amount of forgetting appeared between the 3- and 7-day retention intervals at any age. The second experiment showed that forgetting in juveniles was alleviated by a single training trial administered just prior to the probe trial. These results are discussed in terms of ontogenetic differences in memory processing and measurement issues pertinent to the Morris water maze test procedure.