To determine the prognostic value of the treadmill exercise test, we evaluated 2842 consecutive patients with chest pain who had both treadmill testing cardiac catheterization. The population was randomly divided into two equal-sized groups and the Cox regression model was used in one to form a treadmill score that was then validated in the other group. The final treadmill score was calculated as follows: exercise time--(5 X ST deviation)--(4 X treadmill angina index). Using this treadmill score, 13% of the patients were found to be at high risk; 53%, at moderate risk; and 34%, at low risk. The treadmill score added independent prognostic information to that provided by clinical data, coronary anatomy, and left ventricular ejection fraction: patients with three-vessel disease with a score of -11 or less had a 5-year survival rate of 67%, and those with a score of +7 or more had a 5-year survival rate of 93%. The treadmill score was useful for stratifying prognosis in patients with suspected coronary artery disease who were referred to us for catheterization, and may provide a useful adjunct to clinical decision making in the larger population of patients being evaluated for chest pain.