Twenty-four subjects, each having a valid Virginia driver's license (males and females of ages from 20 to 73), read aloud words presented on two displays while driving a simulated vehicle in night-time conditions. The words, emulating printed legends found on automobile instrument panels, were presented in different hues (eight levels), brightnesses (two levels), character sizes (four levels), and word complexities (two levels). The brightness levels had been subjectively determined in a preliminary experiment by subjects representative of the older and younger segments of the driver population. Each of two groups of drivers determined one brightness level that was subjectively equal among the eight hues. For each word presented, six reading and driving performance measures were taken. Also, subjective attractiveness, subjective comfort, and subjective ease of readability of each hue by brightness treatment combination, were measured. Globally, the results tend to indicate that color of illumination per se had a reliable effect on subjective preferences but a negligible effect on reading and driving performance. Brightness had an impact on performance only with the smaller character sizes. For the larger sizes, brightness level as selected by the subjects (in the preliminary experiment) had a negligible effect on performance. Character size had marked effects on both performance and subjective preferences. The two smaller character sizes tested yielded significant performance decrements for older drivers while the two larger sizes yielded best performance and were better accepted by all subjects. Word complexity did show a significant effect on glance time at the displays with all character sizes. The results of the experiment were transformed into a set of guidelines for use in design of automobile instrument panels. A number of recommendations for future research are also included. / Ph. D.