Background. Considerable evidence exists that some reading-disabled children have disordered visual processing, specifically in the fast processing magnocellular (M) pathway. Methods. The extent that varying luminance and temporal frequency affect amplitude and latency of visual evoked potentials (VEPs) in normally achieving and reading-disabled children grades 4 to 6 was measured. Each group consisted of approximately 30 subjects. Monocular and binocular single channel VEPs were recorded using a sinusoidal checkerboard pattern of spatial frequency 14 min arc at 3 different temporal frequencies (1, 4, and 8 Hz), and an 8 Hz flicker fusion stimulus. Stimuli were presented under high and low luminance conditions. The peak of the major positive wave component (P1Oo) of each waveform and the trough of the previous major negative wave component were identified, and the peak to trough amplitude was measured. Results. Statistical analysis of the VEP amplitudes and latencies in response to different experimental conditions was performed using a repeated measure analysis of variance (MANOVA). VEP amplitudes were significantly higher for normal readers across all conditions. Within all subjects, significant effects were found for monocular vs. binocular viewing, temporal frequencies, and high vs. low luminance. Similar analysis of latencies revealed no significant differences. Conclusions. The presence of a weaker VEP response in reading-disabled children suggests a deficit early in visual processing. The significant difference in VEP amplitudes between the two reading groups provides an objective measure of a deficit in the M pathway that has been implicated in this condition. Whether serial VEP recordings might help to assess the effects of optometric therapy by providing an independent index of therapeutic efficiency is of special interest.