The purpose of this study was to compare the coronary heart disease (CHD) probability estimates of a group of sedentary males involved in an exercise stress test program from 1968 through 1974 with those of a comparison group of sedentary males not involved in the program. The program was designed to evaluate cardiopulmonary function and improve cardiopulmonary efficiency through exercise programs on the basis of performance feedback from periodic exercise stress test evaluation. Data were collected annually, and CHD risk scores were developed using the following variables: age, serum cholesterol, systolic blood pressure, hemoglobin, relative body weight, cigarette smoking, and ECG patterns. Risk scores for the two groups as a whole did not differ significantly upon entry into the program, but for each year of participation in the exercise stress test program, scores for the comparison group were significantly greater than those in the test group. Conclusions drawn from the study are (1) probability estimates for quantifying the likelihood of CHD development are beneficial in identifying individuals for examination or treatment and for observing difference in risk among groups of people; and (2) sedentary individuals who participate on a regular basis in an exercise program have a significantly lower probability of developing CHD than individuals who do not participate in such a program. (MB)