This study was designed to investigate the relationship of eye movements of grade five pupils to variations in readability levels. A further purpose was to assess the effects of these variations on the eye movements of pupils of poor, average, and good reading ability. Twenty pupils were selected at random for each of three groupings of poor, average, and good readers from a total of 213 grade five students. Each subject's independent, instructional, and frustration level was determined using the Botel Reading Inventory A. Eye movements were recorded for each subject at each of the three levels using E.D.L's Biometric Reading Eye II and passages from the Reading Eye Test Selections. Number of fixations and regression, average duration of fixation, average reading rate, and comprehension percentage were computed for each recording. Analysis of variance was used to determine the significance of both inter-group and intra-group variations and post-hoc comparisons were made using Student Newman-Keuls test. The fiver per cent level of significance was set as the acceptable level for significant difference. Analysis of the results revealed significant differences between poor, average, and good readers in the number of fixations, regressions, average span of recognition, and average reading rate when reading at their independent, instructional, and frustration levels. Poor readers also exhibited a significantly longer average duration of fixation than either average or good readers at their independent, instructional, and frustration levels but no significant difference in average duration of fixation was found between average and good readers at these levels. No significant differences were found in the number of fixations, number of regressions, and average span of recognition of grade five students reading at their independent or instructional levels. Significant differences were found in all eye movement components, number of fixations, number of number of regressions, average span of recognition, average duration of fixation, and average reading rate, of grade five students reading at their independent or frustration levels. Graphs of the mean values for the eye movements of average and good readers indicated a linear trend from independent to instructional to frustration levels for average span of recognition and average duration of fixation. The performance of poor readers for average span of recognition and average duration of fixation did not follow the same pattern as that for average and good readers but, rather, suggested a significant relationship between appropriateness of material (instructional level) and eye movement behavior which warrants further investigation. Further support for this relationship was provided by the superior performance of poor and average readers in terms of the number of fixations and regressions exhibited at their instructional level as opposed to the independent or frustration levels. / 1976