THE HYPOTHESIS THAT CHILDREN WHO ARE TAUGHT TO HEAR AND DESIGNATE SEPARATE SOUNDS IN SPOKEN WORDS WILL ACHIEVE GREATER SUCCESS IN LEARNING TO ANALYZE PRINTED WORDS WAS TESTED. THE SUBJECTS WERE 90 KINDERGARTEN CHILDREN, PREDOMINATELY MEXICAN-AMERICANS AND NEGROES. CHILDREN WERE RANDOMLY ASSIGNED TO ONE OF THREE TREATMENTS, EACH OF 3-WEEKS DURATION -- AN AUTOINSTRUCTIONAL PROGRAM IN BEGINNING READING, AN AUTOINSTRUCTIONAL AUDITORY TRAINING PROGRAM THAT USED NO VISUAL STIMULI, AND NONINSTRUCTION IN READING AND AUDITORY DISCRIMINATION. AFTER THE THREE EXPERIMENTAL TREATMENTS, ALL THE CHILDREN RECEIVED PROGRAMED LESSONS IN READING IDENTICAL TO THOSE IN THE FIRST TREATMENT ABOVE. ACHIEVEMENT WAS MEASURED BY ERRORS DURING INSTRUCTION IN READING, SCORES EARNED ON A TEST OF WORD ANALYSIS SKILLS, AND SCORES EARNED ON AN AUDITORY DISCRIMINATION TEST. IT WAS FOUND THAT THE AUDITORY TREATMENT RESULTED IN A LOWER READING ERROR RATE AND MORE SUCCESSFUL PERFORMANCES IN BOTH THE TASK OF WORD ANALYSIS AND AUDITORY DISCRIMINATION. THE EXCLUSIVELY AUDITORY PROGRAM FOLLOWED BY INSTRUCTION IN READING MORE EFFECTIVELY IMPROVED SKILLS FOR ANALYZING PRINTED WORDS. AUDITORY TRAINING ON SELECTED PHONEMES HELPED LEARNERS ASSOCIATE GRAPHEMES AND PHONEMES EVEN WHEN THE PHONEMES WERE DIFFERENT FROM THOSE IN TRAINING. AUDITORY TRAINING WAS ESPECIALLY BENEFICIAL TO BOYS AND CHILDREN OF LOWER INTELLIGENCE. REFERENCES, APPENDIXES, AND EXAMPLES OF TESTS AND LESSONS USED ARE INCLUDED. (MC)