This study investigated the value of pharmacologic stress echocardiography for risk stratification of patients > or =65 years of age. The study cohort consisted of 2,160 patients > or =65 years of age (1,257 men, mean +/- SD 71 +/- 5 years of age) undergoing dipyridamole (n = 1,521) or dobutamine (n = 639) stress echocardiography for evaluation of known (n = 913) or suspected (n = 1,247) coronary artery disease. Of 2,160 patients, 753 (35%) had a normal test result, whereas 772 (36%) showed a myocardial ischemic pattern and 635 (29%) a scar pattern. During a median follow-up of 26 months, 241 deaths and 87 nonfatal myocardial infarctions occurred. Patients (n = 568) undergoing revascularization were censored. Of 16 analyzed variables, age (hazard ratio [HR] 1.07 per unit increment), wall motion score index at rest (HR 2.63 per unit increment), ischemia at stress echocardiography (HR 1.81), and diabetes (HR 1.57) were multivariable predictors of death, whereas age (HR 1.06 per unit increment), ischemia at stress echocardiography (HR 2.60), wall motion score index at rest (HR 1.98 per unit increment), scar pattern (HR 1.99), and diabetes (HR 1.48) were multivariable predictors of death or myocardial infarction. Using an interactive stepwise procedure, stress echocardiography showed incremental prognostic value over clinical and echocardiographic data at rest, which decreased with increasing age. In addition, the annual hard event rate associated with a normal test result progressively increased with age. In conclusion, pharmacologic stress echocardiography provides useful prognostic information in patients > or =65 years of age. However, its prognostic value decreases with increasing age.