During a four-year period, 6,650 persons were enrolled in the exercise testing registry of the Seattle Heart Watch Study. Risk factors and maximal exercise responses were assessed. The purpose of the study was to document the prognostic value of these variables for primary cardiac events in asymptomatic persons and for secondary cardiac events in persons with known heart disease. In asymptomatic persons, age and the combination of risk factors and abnormal exercise responses identified individuals at increased risk. In persons with established coronary heart disease, age and conventional risk factors were of little prognostic value. However, findings that included cardiomegaly, poor blood pressure response during exercise, and limited exercise tolerance, all of which suggested poor left ventricular function, were predictive of poor prognosis. Exertional chest pain and abnormal ST depression were not predictive.