In this study we recorded eye movements during reading tasks for five dyslexic children and two controls. The subjects had vertical as well as horizontal eye position monitored with an infrared recording technique and performed a symbol-simulated reading task. We demonstrated the importance of both the symbol simulation and the vertical monitoring in evaluating the eye-movement patterns in dyslexics. Without the symbols, motor and comprehension defects cannot be distinguished. Without vertical eye position information, the horizontal eye-movement pattern is not properly interpretable.