Rote learning and retention performance was studied as a function of method used in original learning and as a function of intellectual level. Sixty educable mentally retarded and 60 mentally normal junior high school students were randomly selected and assigned to one of three treatment groups, each learning to a different criterion, for each intellectual category in order to learn a paired associate task. Retention was assessed by immediate recall scores, 24 hour recall scores, and relearning scores following the 24 hour interval. A 2x3 complete factorial analysis of covariance was performed for the following dependent variables: original learning; relearning; and 24 hour recall. Immediate recall was assessed utilizing a 2x3 complete analysis of the variance procedure. The results of the investigation indicated inferior learning performance and a 24 hour retention deficit for retarded subjects; and amelioration by overlearning of retention deficits in the retarded subjects. In addition, the results in the area of rote learning and retention comparing mentally retarded and normal subjects were found to be method dependent. (Author/JD)