Two conflicting hypotheses were investigated in this study. The first is based on an interference theory of forgetting and assumes that forgetting is largely a function of proactive interference, i.e., a result of interference by previously learned material. The second is based on the improvement of memory due to basic mechanisms that may operate to produce marked increases in retentive abilities over the first few years of a child's life. The subjects were 45 kindergartners and 45 third graders randomly assigned to three experimental groups. The experimental conditions varied the amount of retention time between the original learning task and the retention tests. The retention intervals were six minutes, 48 hours, and four weeks. The learning task was a list of four paired associates. In each pair, the response term was a high frequency, one-syllable English noun and the stimulus term was the Latin equivalent. The learning task and the retention tests were all conducted orally. The number of items correct on the three retention tests served as the dependent measure for analysis. The results indicated that when the degree of original learning is equated, children as young as five years of age can remember materials as well as third-graders over a retention interval as long as four weeks. (WR)