THE PURPOSE OF THIS STUDY WAS TO INVESTIGATE A TESTING METHOD FOR OBTAINING OBJECTIVE, QUANTIFIED, PREDICTIVE MEASURES OF EDUCABILITY FOR MENTAL RETARDATES. THE METHOD WAS DESIGNED TO MEASURE ABILITIES ON WHICH EDUCATION DEPENDS--NAMELY, THE ABILITY TO LEARN UNDER TRAINING, TO TRANSFER THE TRAINING, AND TO RETAIN THE LEARNING. TWO SAMPLES WERE USED--A PILOT GROUP OF 115 AND A FINAL GROUP OF 114. SUBJECTS IN THE FINAL GROUP HAD A MEAN IQ OF 36 AND AN AVERAGE AGE OF 95 MONTHS. THE FINAL TEST BATTERY WAS ADMINISTERED INDIVIDUALLY TO TEACH EACH SUBJECT AND INCLUDED A TEST OF IMITATIVE ABILITIES AND MEMORY, A SIZE DISCRIMINATION TEST, A SHAPE DISCRIMINATION TEST, AND TWO TESTS OF BRIGHTNESS DISCRIMINATION. ALL OF THE SUBJECTS RECEIVED THE FIRST FOUR TEST ADMINISTRATIONS WHICH YIELDED PRETRAINING, POST-TRAINING, TRANSFER, AND RETENTION SCORES RESPECTIVELY. OF THE TOTAL SAMPLE, 102 TOOK THE FIFTH TEST SERIES, BASED ON RETENTION, AND 36 WERE AVAILABLE FOR THE LAST TESTS, REPETITIONS OF THE FIRST 4 TESTS. INTELLIGENCE EXERTED A GREATER SIGNIFICANT EFFECT ON THE SCORES OF EVERY TEST THAN THE FACTORS OF AGE, TIME SPENT IN SCHOOL, SEX, OR MEDICAL DIAGNOSTIC CATEGORY. THE RELATIONSHIP BETWEEN LEARNING PERFORMANCE AND INTELLIGENCE LEVEL WAS SIGNIFICANTLY GREATER ON EVERY TEST THAN THAT BETWEEN INTELLIGENCE LEVEL AND INITIAL PERFORMANCE. THE ANCILLARY FINDINGS SUGGEST THE APPLICABILITY OF THE MATERIALS AND TRAINING METHODS DEVELOPED FOR THE TESTS TO CURRICULAR MATERIALS AND CLASSROOM METHODS, AND THE SUITABILITY OF TESTING USED FOR OTHER GROUPS OF CHILDREN. (JH)