A 10 year prospective community practice study in Seattle of risk of primary morbidity (defined by hospital admission) and mortality due to coronary heart disease in 3,611 men and 547 women initially free of clinical manifestations of this disease revealed a crude incidence of 202 coronary heart disease events, or 4.9% in 6.1 +/- 2.6 years of follow-up. The case fatality rate was 16.8%. Stratification by clinical classification of asymptomatic healthy persons versus patients with atypical chest pain syndrome (not angina pectoris) and hypertension (as classified by physicians) showed an incidence rate of primary events due to coronary heart disease of 2.9, 5.5 (not significant) and 10.0% (p less than 0.001), respectively. Identification of conventional risk factors is known to be important for risk assessment. However, the presence of any conventional risk factor, in conjunction with two or more selected maximal exercise predictors (which vary with the clinical classification) at enrollment, substantially increased the cumulative 6 year incidence rate to 24.3, 15.5 and 33.3% in asymptomatic healthy men, patients with atypical chest pain syndrome and hypertensive patients, respectively. Observation of the exercise predictors in the absence of conventional risk factors increased the risk much less, suggesting that the use of maximal exercise testing for risk assessment in those with no clinical manifestations of disease might be limited to persons with one or more conventional risk factors.