TO COMPARE CONVENTIONAL AND PROGRAMED INSTRUCTION ON TEACHING AUDITORY DISCRIMINATION OF THE "S" PHONEME, 20 LESSONS CONTAINING STIMULUS AND REPEAT ITEMS WERE RECORDED ON A COMMERCIAL MULTI- CHANNEL DEVICE THAT ALLOWS CONTINUOUS ADJUSTMENT ON THE BASIS OF RESPONSE. THE 40 GRADES TWO-FOUR PARTICIPANTS, MATCHED ON AGE, SEX, AND IQ, WERE GIVEN THREE ARTICULATION AND DISCRIMINATION ACHIEVEMENT TESTS BEFORE, IMMEDIATELY AFTER, AND ONE MONTH AFTER INSTRUCTION. EXPERIMENTAL SUBJECTS WORKED SOLELY ON THE EQUIPMENT, WHEREAS CONTROL SUBJECTS RESPONDED VERBALLY TO DISCRIMINATION TASKS AND WERE NOT GIVEN POSITIVE REINFORCEMENT FOR CORRECT RESPONSES. BOTH GROUPS SHOWED EQUALLY SIGNIFICANT LEARNING IN, AND RETENTION OF, PHONEME AND GENERAL DISCRIMINATION, AND PHONEME AND GENERAL ARTICULATION, ALTHOUGH NO SPECIFIC INSTRUCTION WAS GIVEN FOR THE LAST THREE SKILLS. TIME TO COMPLETION WAS LONGER FOR THE EXPERIMENTAL GROUP. IN VIEW OF THE CRITICAL SHORTAGE OF QUALIFIED SPEECH PATHOLOGISTS IN THE PUBLIC SCHOOLS, IT WAS CONCLUDED THAT ADDITION OF AUTOMATED INSTRUCTION IN SPEECH PATHOLOGY IS DESIRABLE AND PRACTICAL. (LH)