Driving is a dynamic task. Hence, for a given message, readability depends on legibility distance and reading time. In laboratory simulation and outdoor tests, the variables of pixel shape, letter width/height ratio, font and letter separation were studied. Nearly 100 subjects took part in the laboratory simulation using monochrome slides. Half the subjects assessed legibility distance while the remainder determined reading time. The method of limits was used to determine the thresholds for legibility distance and reading time for nearly 50 different formats of matrix sign. For legibility distance, the thresholds found in the laboratory were validated using real signs in an outdoor situation. However, it was not possible to validate reading times in a similar manner due to lack of availability of sufficiently large full scale signs. Variations in format had a marked effect on legibility distance, but a lesser impact on reading time. In general, letters with any regular shaped pixel, a width/height ratio approaching 1.0, an upper case font and a letter separation of two pixels were found to be most effective.