The purpose of this study was to investigate the longitudinal effects of i/t/a and T.O. instruction on students' reading, spelling, and language abilities at the end of third grade. It was primarily concerned with determining whether the effects of early reading instruction, which were not statistically evident at the end of second grade, became significant at the end of third grade. A summary of the analyses of variance and covariance computed for the orthography used (i/t/a or T.O.) suggested that i/t/a produces significantly better reading achievement in word study skills and word recognition and that i/t/a children spell as well as T.O. children by the end of third grade. From the data collected the author concluded that no significant advantage accrues to a total group from beginning reading instruction on a universal basis prior to first grade and that the Initial Teaching Alphabet proved to be a superior medium of instruction regardless of the time at which instruction began. However, i/t/a superiorities were not consistently evident in the area of comprehension. References are included. (NH)