A problem in divided visual field studies which use event-related potentials as a dependent measure is the large number of horizontal eye movements participants make during experimental trials. Past attention research suggests that eye movements to lateralized targets should be significantly reduced using a dynamic, offset mask, causing a reduction in attentional capture. The current study attempted to replicate past divided visual field language studies using offset masking procedures. Using a basic offset procedure, eye movements were not reduced in Experiment 1. Experiment 3, however, did see a significant reduction in eye movements using a dynamic offset masking procedure developed in Experiment 2. Low accuracy rates were a concern throughout. In conclusion, horizontal eye movements can be reduced with a dynamic offset procedure but the low accuracy rates and the inconsistent behavioral findings throughout the study do not support using this technique.