The unstable angina pectoris (UAP) classification proposed by Braunwald in 1989, although often used, has never been validated in a large, prospective multicenter study in which all subgroups of patients were included. Patients with UAP or non-ST-elevation myocardial infarction (NSTEMI) were enrolled in the Thrombolysis In Myocardial Ischemia III Registry and classified according to the Braunwald classification for UAP. Clinical end points were compared at 6 weeks and 1 year. Of 3,318 patients, those with primary UAP had lower rates of recurrent myocardial infarction (MI) or death when compared with patients with secondary UAP and post-MI UAP at 6 weeks (4.1% vs 6.4% vs 13.4%, respectively; p <0.001) and 1 year (9.7% vs 16.7% vs 19.7%; p <0.001). Recurrent ischemia at 6 weeks followed the same gradient (13.2% vs 18.5% vs 20.8%; p <0.001). Patients with secondary UAP had similar extent of disease at angiography as primary UAP. Patients with nonresting UAP had lower rates of death or MI than patients with UAP at rest (3.0% vs 5.6%, p = 0.011 at 6 weeks, and 8.2% vs 12.5%, p = 0.004 at 1 year). Patients with ST-segment deviation and those who had received prior antianginal medical treatment also had worse outcomes. Thus, the Braunwald classification of UAP predicts prognosis with secondary UAP, post-MI UAP, and patients with pain at rest who have a higher risk for death or recurrent cardiac events. Given their high risk for adverse events, patients with secondary UAP should be treated aggressively.