Eighteen retarded Ss (mean IQ 50 and mean age 14 years) and 18 normal Ss (mean IQ 100 and mean age 7 years) participated in a study to isolate variables that differentially control discrimination learning and retention processes, and to evaluate contrasting theories on discrimination learning and menory processes of retarded and normal children. The experimental design contained the following six independent variables: IQ groups (retardates vs. normal Ss), retention intervals (seconds vs. hours), kinds of test problems (new vs. old), sources of interference (proactive vs. retroactive), types of interfering discrimination (new vs. old), and Ss. The dependent variable in all conditions was the S's response (on a discrimination test trial) expressed in terms of either percentage of correct responses or retention loss score in each level of the factorial experiment. Acquisition data showed that retarded Ss were slightly but significantly slower learners than normal Ss and required more trials to reach criterion on interference free discrimination problems. Retention data showed that retarded Ss had only a slight overall retention deficit compared to normal Ss, and that both groups displayed greatest retention loss on new test problems, long retention intervals, and interference conditions (retroactive conditions for the retarded Ss and proactive conditions for the normal Ss). The results indicated a dual memory process of retarded and normal children and therefore supported N. Ellis's 1970 Primary and Secondary Memory Theory and the data of D. Zeaman's 1973 Attention-Retention Theory. (Author/MC)