The comparative use of grapho/phonic and syntactic/semantic cues for ongoing word recognition by readers of varying grade levels was examined in both oral and silent reading modes. Twenty-four subjects in each of four grades (2, 4, 6, and 8) read grade-appropriate passages orally and silently. Target passages each had an orthographically altered word embedded in the text. Results included a marked difference in cue system implementation in oral and silent reading modes. A greater number of altered words were identified in the oral reading mode. This difference was statistically significant in three of the four grade levels: Grade 4, F(1, 23) = 19.58, p<.001; in Grade 6, F(1, 23) = 9.38, p<.005; and in Grade 8, F(1, 23) = 6.18, p<.02. These results cast suspicion on diagnostic and instructional practices which assume similar word recognition processing in both reading modes.