A study tested the instructional hypothesis that variation in instructional methods in the initial stages of formal reading development will differentially develop subsystems for attaining comprehension. The 91 kindergarten and first grade students in the study received their usual reading instruction plus supplementary instruction in one of four specified treatment conditions that emphasized development of a particular linguistic subsystem (graphophonemics, morphophonemics, semantics, and syntax). The Metropolitan Reading Instructional Tests (MRIT) and specially developed criterion referenced tests (CRT) were used to measure the dependent variables. Analyses of variance indicated that supplementary instruction had a differential but statistically significant effect in the expected direction on all four of the CRT subtests and on one of the six MRIT subtests. There were no significant differences between treatment groups on reading comprehension. These results supported the hypothesis that variation in instructional methods leads to differential development of the subsystems underlying achievement in reading comprehension. (RL)