The purpose of the study was to determine whether serial graded exercise test (GXT) responses, known to have prognostic importance, in the coronary artery disease population generally, can differentiate certain cardiovascular morbidity and mortality endpoints among physically active patients. Data were obtained in 1989 at three community-based cardiac rehabilitation programs from three serial GXT results in 23 paired cardiac event (CE) and non-event (NE) patients. A cardiac event was defined as having an myocardial infarction (MI), percutaneous translumial angioplasty (PTCA), coronary artery bypass grafting (CABG), and cardiac death. Comparisons were made just prior to each CE patients event. Statistical analyses were done on four exercise test variables: ST-segment depression (ST80) at peak exercise, heart rate adjusted 5Tsegment depression (ST/HR slope), exercise induced ventricular dysrhythmias (PVC), and exercise induced chest pain. Chi-square analyses showed no significant changes across the serial tests for any of the variables examined in either the CE or NE groups (P)0.05). Nevertheless, the CE group had significantly more 8Tsegment depression, ST/HR slope, and exercise chest pain in all GXT intervals in comparison to the NE group. These findings do not support the use of serial exercise GXTs for prognostic information for the physically active cardiac rehabilitation maintenance patients. / Master of Science