Traditional approaches to the evaluation of a computer interface typically uses methods such as cognitive walk through, GOMs, task analysis, user's performance and self-report. These are effective approaches; however, they are more subjective than objective. Eye-movement analysis provides a unique and useful strategy for evaluating a computer interface. Additionally, it allows a quantitative approach to understand how users interact with the computer. In this paper, background color, icon size and number of icon presentation were tested on participants with Agerelated Macular Degeneration. The experimental design is a matching to sample task where participants were presented with a sample icon and the task is to match the sample icon to a set of icons. Eye-movement data were recorded through video output and digitized for analysis. Results of the study showed that icon size and the number of icon presentation have a significant effect on the time participants took to search and identify the target icon.