Of the subjective rating scales utilized in workload research the Cooper-Harper scale is the most popular. This decision tree rating scale and modifications of the scale have provided sensitive measurements by discriminating among various mental workload levels. This research is an investigation into the characteristics of a Modified Cooper-Harper rating scale (MCH) developed by Wierwille and Casali (1983). Six rating scale designs emphasizing major characteristics which might cause the MCH scale to be a sensitive measure of mental workload were used in this study. The aims of the research were to discover modifications of the MCH might make it more sensitive. A communications task developed by Casali and Wierwille (1983) was manipulated to present 36 subject pilots, both private and student, with three communications loading levels. The pilots were distributed into the six rating scales by experience level. Six different experience levels were represented in each of the rating scale groupings. Using the communications loading, the presence of a decision tree in the scales appeared to improve the scale's ability to discriminate among loading levels. The expansion of the MCH scale to 15 categories decreased the sensitivity of the MCH rating scale. The standard 10-point MCH rating scale was the most consistent of the six rating scales and attained a high ability to discriminate among loading levels. Finally, a companion mediational task study by Rieger (1983) using the same six rating scales resulted in substantially different results, suggesting that subjective rating scales are extremely task dependent. / M.S.