AN EXPERIMENTAL GROUP OF 461 CHILDREN WAS TAUGHT TO READ WITH THE INITIAL TEACHING ALPHABET (I/T/A) BEGINNING IN KINDERGARTEN AND WAS READING TRADITIONAL ORTHOGRAPHY (T/O) MATERIALS BY THE END OF THE SECOND GRADE. THE CONTROL GROUP INCLUDED 462 CHILDREN WHO WERE TAUGHT TRADITIONAL ORTHOGRAPHY IN A PROGRAM CENTERED AROUND CONVENTIONAL BASAL READING SERIES. AT THE END OF THE SECOND GRADE, BOTH GROUPS WERE ADMINISTERED THE PARAGRAPH MEANING, WORD MEANING, AND SPELLING SECTIONS OF THE STANFORD ACHIEVEMENT TEST, PRIMARY, FORM K. A STATISTICAL COMPARISON OF SCORES ON EACH TEST SECTION WAS MADE. THIS WAS THE SECOND STEP IN A LONGITUDINAL INVESTIGATION OF THE EFFECTS OF I/T/A. THE FIRST COMPARISON WAS MADE AT THE END OF GRADE ONE. AFTER FIRST GRADE, THE I/T/A GROUP DID SIGNIFICANTLY BETTER ONLY IN SPELLING WHEN EITHER I/T/A OR T/O ANSWERS WERE ACCEPTABLE. AT THE END OF GRADE TWO, THERE WERE NO SIGNIFICANT DIFFERENCES IN THE THREE AREAS. THE I/T/A METHOD APPEARED TO BE BEST SUITED FOR ABOVE AVERAGE CHILDREN. ALTHOUGH STAFF OPINION WAS NOT UNANIMOUS, IT WAS RECOMMENDED THAT THE PROGRAM BE DISCONTINUED. TWELVE REASONS FOR THIS RECOMMENDATION ARE GIVEN. STATISTICS ARE SUMMARIZED IN FOUR TABLES. (RH)