The American College of Sports Medicine (ACSM) has published exercise guidelines identifying individuals who should have an exercise test prior to clearance for exercise participation and whether a physician should supervise the test. These age and health status criteria (apparently healthy, higher risk, and diseased) were developed using clinical judgement and opinion rather than empirical data. Thus, there is a need to validate the recommendations with actual data. We studied the results of 24,332 maximal treadmill tests in men (n = 18,076) and women (n = 5,626) as they associated with age and baseline health status. Commonly accepted criteria for abnormal exercise tests were used (i.e., 1 mm ST segment depression at 0.08 s, systolic blood pressure drop with exercise, complex ventricular ectopy, etc.). There were 895 and 183 abnormal exercise tests in men and women, respectively. Men and women who were apparently healthy had lower abnormality rates per 1,000 tests than those considered to be at higher risk for coronary heart disease and those who had preexisting disease. Further, when those who were at higher risk were considered, those men with only one risk factor had significantly lower abnormality rates than did men with more than one risk factor (95% confidence intervals (CI) per 1,000 tests: 1 risk factor = 36.1-46.4; greater than 1 risk factor = 47.5-62.5). Abnormality rates in women with 1 risk factor were also lower than those in their peers with greater than 1 risk factor, but not statistically so (95% CI per 1,000 tests: 1 risk factor = 24.9-43.0; greater than 1 risk factor = 25.3-54.4).(ABSTRACT TRUNCATED AT 250 WORDS)