The characteristics of visually guided saccadic eye movements were compared in 12 normal adult subjects and 12 test subjects of normal intelligence, but with problems in reading and writing. All subjects were examined psychometrically for different cognitive abilities, and for their reading and writing capabilities. The anamnestic reports about their reading and writing problems earlier in their lives were analysed. Based on scores of the writing and reading tests, the test group was subdivided into two subgroups: A and B (group A reached medium, group B very low scores in both tests). Five different non-cognitive eye movement tasks were applied: two single tasks (gap and overlap) requiring single saccades from a fixation point to a peripherally appearing target and three sequential tasks (overlap, synchronous, and simultaneous) requiring sequences of saccades to four equally spaced targets presented sequentially to the right side from an initial fixation point. Many parameters of the subjects' eye movement performance were determined and their mean values were calculated for each subject. The Student t-test revealed that the eye movement data of the two test groups deviated differently from the data of the control group. Group B had the largest deviation of the eye movement parameter from the control group. The differentiating parameters were the consistency of target acquisition, the saccadic reaction time, and the number of anticipatory responses in the single target tasks. In the sequential tasks these differences were in the amplitude, the number of saccades, and the fixation durations. The incidence of regressive saccades did not discriminate between test subjects and controls. The possible underlying deficits in the attentional control over the saccadic system and their implications for reading are discussed.