A 2-YEAR LONGITUDINAL STUDY WAS CONDUCTED TO EVALUATE THE "EARLY-TO-READ I/T/A PROGRAM." THIRTY-FOUR PUPILS RAMDOMLY ASSIGNED TO ONE FIRST GRADE WERE TAUGHT USING THE "EARLY-TO-READ I/T/A SERIES," AND 26 PUPILS RANDOMLY ASSIGNED TO ANOTHER FIRST GRADE WERE TAUGHT USING THE "GINN BASIC READERS" AND TRADITIONAL ORTHOGRAPHY (TO). A SUBCONTROL GROUP OF 86 PUPILS WAS SELECTED RANDOMLY FROM THE REMAINING FIRST GRADES AND RECEIVED ONLY INITIAL READINESS TESTING AND FINAL READING ACHIEVEMENT TESTING AT THE END OF GRADES 1 AND 2. TESTS ADMINISTERED WERE THE PREREADING TEST ACCOMPANYING THE SHELDON READING SERIES, A LETTER KNOWLEDGE TEST OF THE ALPHABET, THE WECHSLER INTELLIGENCE SCALE FOR CHILDREN (WISC), THE GRAY ORAL READING TEST, THE STANDARD READING INVENTORY, AND THE STANFORD ACHIEVEMENT TESTS. THERE WERE NO DIFFERENCES AMONG THE GROUPS IN READINESS TO READ. THERE WAS NO SIGNIFICANT DIFFERENCE IN OVERALL READING ACHIEVEMENT BETWEEN THE EXPERIMENTAL AND MAIN CONTROL GROUP AT THE END OF FIRST OR SECOND GRADE WHEN THE READING WAS DONE IN TO. THERE WERE FAIRLY SIGNIFICANT DIFFERENCES BETWEEN BOTH THE EXPERIMENTAL AND SUBCONTROL GROUPS, AND BETWEEN THE CONTROL AND SUBCONTROL GROUPS ON MOST MEASURES OF READING ACHIEVEMENT AT THE END OF FIRST AND SECOND GRADES. THE EXPERIMENTAL GROUP CONSISTENTLY ACHIEVED THE HIGHEST SCORES, AND THE SUBCONTROL GROUP CONSISTENTLY ACHIEVED LOWEST SCORES. THESE AND OTHER RESULTS SEEMED TO INDICATE THAT GOOD TEACHING UNDER TRADITIONAL ORTHOGRAPHY WAS PREFERABLE TO CHANGING THE ALPHABET. (BK)