Background Certain characteristics of unstable angina have been associated with worse clinical outcomes after percutaneous revascularization procedures. Methods And Results We compared outcomes of patients with (n = 690) and those without (n = 320) unstable angina in the Coronary Angioplasty Versus Excisional Atherectomy Trial (CAVEAT-I) by unstable angina subgroup: rest chest pain, postinfarction chest pain, chest pain with ischemic electrocardiographic changes, chest pain with recent acceleration, and no chest pain. Regression models were constructed to predict in-hospital and 6-month composite end point death, infarction, bypass surgery, percutaneous revascularization, and abrupt closure (in-hospital) or restenosis (6 months) for each subgroup. Only chest pain with electrocardiographic changes predicted the composite in-hospital outcome (24% vs 17% with no chest pain, P =.0374.) This subgroup also had a greater acute gain, more late loss, and more restenosis than patients in the other subgroups. Rest chest pain carried a higher incidence of the composite 6-month outcome (39.9% vs 29% with no chest pain, P =.0472). For all unstable angina categories, atherectomy was associated with worse overall outcomes than angioplasty. Conclusions Patients with unstable angina have more complications of percutaneous revascularization than patients without unstable angina, but event rates vary by anginal subgroup. The clinical presentation may help to identify unstable angina patients at particularly high risk for adverse outcomes. (Am Heart J 1999;137:227-33.)