A STUDY WAS CONDUCTED TO DELINEATE HOW PERCEPTION OCCURS DURING ORAL READING. FROM AN ANALYSIS OF CLASSICAL AND MODERN RESEARCH, A HEURISTIC MODEL WAS CONSTRUCTED WHICH DELINEATED THE DIRECTLY INTERACTING SYSTEMS POSTULATED AS FUNCTIONING DURING ORAL READING. THE MODEL AS OUTLINED WAS DIFFERENTIATED LOGICALLY INTO THREE MAJOR PROCESSING FUNCTIONS--SENSORY, RECOGNITIONAL, AND RESPONSE. DATA WERE DERIVED FROM 2,465 EYE-VOICE SPAN PAIRINGS FURNISHED BY EIGHT SUBJECTS READING THREE PASSAGES OF VARYING DIFFICULTY. THE SYNCHRONIZATION OF RECORDINGS USING THE GILBERT EYE-MOVEMENT CAMERA AND TAPED ORAL READING SHOWED RELATIVELY CONSTANT EYE-VOICE SPAN. CORRELATIONS BETWEEN MEASURES OF SMOOTH READING AND THE CONSTANCY OF TEMPORAL EYE-VOICE SPAN WERE HIGH. THE ESTIMATED TIME ELAPSING BETWEEN EYE AND VOICE DURING SMOOTH READING WAS VALIDATED. REESTABLISHING A STEADY STATE FOLLOWING AN INTERRUPTION WAS NOT STATISTICALLY STABLE. THE USE OF MULTIPLE FIXATIONS, REGRESSIONS, AND PROLONGED FIXATION PAUSES WAS SUBSTANTIATED, AND SIGNIFICANT RELATIONSHIPS AMONG EYE-MOVEMENT PATTERNS WERE FOUND. THE NEED TO TEMPORALLY BALANCE INPUT AND OUTPUT SYSTEMS AND THE NEED TO CORRECT EYE-FUNCTIONING IN ORDER TO REESTABLISH TEMPORAL BALANCE WERE EVIDENT. THE MODEL WAS DECLARED INCOMPLETE, AND, ALTHOUGH IT MAY ADD TO CUMULATIVE KNOWLEDGE OF READING PHENOMENA, HAS LITTLE TO OFFER THE TEACHER AT PRESENT. THIS PAPER WAS PRESENTED AT THE INTERNATIONAL READING ASSN. CONF., (SEATTLE, MAY 4-6, 1967). (MC)