Two hundred fifty-one patients (195 male and 56 females) referred for evaluation of chest pain were studied by multistage submaximal stress testing and selective coronary arteriography. In men with positive exercise tests the incidence rate of true positive exercise test results--that is, positive tests associated with 75 percent of greater coronary stenosis--was 89 percent in contrast to a 33 percent incidence rate of true positive exercise test results in women. The incidence rate of false positive excercise test results--that is, positive tests associated with no coronary stenosis or less than 50 percent stenosis--was 8 percent in men in contrast to 67 percent in women. Conversely, the incidence rate of false negative exercise test results (that is, negative exercise tests associated with 75 percent or greater coronary stenosis) was higher in men (37 percent) than in women (12 percent). It is concluded that in men a positive multistage stress test is useful in predicting the presence of significant coronary artery disease although a negative stress test cannot be relied upon to rule out the presence of significant disease. In women, a positive exercise test is of little value in predicting the presence of significant coronary artery disease, whereas a negative test is quite useful in ruling out the presence of significant disease. New criteria should be developed for stress testing of women.