OBJECTIVE The purpose of this study was to characterize the outcome of coronary angioplasty according to the various presentations of unstable angina pectoris. BACKGROUND Although unstable angina is a mosaic of clinical manifestations, a comprehensive analysis of short- and long-term outcome of coronary angioplasty in subsets of unstable angina is not available. METHODS Data from 15 clinical centers for the 857 patients with unstable angina in the 1985-1986 National Heart, Lung, and Blood Institute percutaneous transluminal coronary angioplasty registry were analyzed. Five-year follow-up was available in > 96.5%. Patients were first classified as those with (679 [79%]) or without (178 [21%]) rest angina. Patients were also allocated to five mutually exclusive categories of decreasing unstable angina severity: postinfarction angina, acute coronary insufficiency, plain rest angina, accelerating angina and new onset angina. RESULTS The group with rest angina had more older patients (p < 0.01) and women (p < 0.001), and a greater proportion had a previous myocardial infarction (p < 0.001) and a left ventricular ejection fraction < or = 50% (p < 0.01) than did the group without rest angina. Angiographic characteristics were nearly the same, whereas procedural characteristics and outcome were the same for both categories. At 5-year follow-up, there was a higher crude mortality rate in patients with than without rest angina (p < 0.05). Resolution into five subsets yielded additional information. Women were more represented only in the acute coronary insufficiency and plain rest angina subsets (p < 0.001). Patients with angina after myocardial infarction had the second shortest history of angina (p < 0.001), the highest percent of smokers (p < 0.01) and, with those with acute coronary insufficiency, the highest incidence of congestive heart failure (p < 0.05) and an ejection fraction < or = 50% (p < 0.001). They had the highest percent of totally occluded arteries, coronary thrombus and collateral blood flow received but also the lowest rate of severe stenoses (p < 0.001 for all). Patients with new onset angina had the highest prevalence of single-vessel disease (p < 0.05), critical and complex stenoses (p < 0.001) and no coronary angioplasty-related deaths. The crude 5-year mortality rate was higher for both postinfarction and acute insufficiency groups (p < 0.05) than for the other subsets. After adjustments for risk factors, no significant differences in adverse event rates remained among the different unstable angina subgroups. CONCLUSIONS Analysis of the diverse clinical presentations of unstable angina supports underlying pathogenetic differences. Coronary angioplasty is safe and effective in all subsets of unstable angina. Long-term survival is good in general but is related to the baseline status of left ventricular function.