The color coding of augmented reality head-up display (AR-HUD) is a crucial factor affecting the cognitive performance of drivers. The work aims to analyze the influence of color matching of the display information in AR-HUD on users, explore the most recognizable and understandable color scheme, and put forward corresponding design suggestions to improve the speed of users in recognizing and understanding information with AR-HUD. Eye movement tracking technology was used to collect eye movement data of subjects watching different color schemes under simulated driving conditions during the day and night, and three fixation indicators were extracted, including the first fixation time, the first fixation duration, and the average fixation duration in the region of interest. Combined with the subjective evaluation of the subjects, the different color schemes of AR-HUD were comprehensively analyzed. In the daytime, the combination of cyan and white or orange can effectively attract the driver's attention, and the recognizability and comprehensibility are higher, the visibility of red is lower, the combination of green and cyan easily causes information confusion and in the strong light environment, the information visibility of yellow is lower. In nighttime environments, yellow with white or cyan can help capture the driver's attention and process information quickly and effectively. The research results can provide reference for the color design of AR-HUD interface.