Baseline features, management, and outcomes of patients who had acute myocardial infarction (AMI) and were excluded from randomized trials of primary percutaneous coronary intervention (PCI) have not been well described. We examined the baseline features and outcomes of patients who had AMI and were excluded due to angiographic ineligibility from a randomized trial of primary PCI. The CADILLAC trial evaluated 4 primary PCI strategies in patients who had AMI without cardiogenic shock. Of 2,681 patients who consented, 599 (22.3%) were subsequently excluded from randomization due to protocol-specified angiographic findings. These patients were enrolled in a formal in-hospital registry and were treated at the discretion of attending physicians. Registry versus randomized patients were older (median age 61.9 vs 59.0 years, p = 0.002), more frequently had 3-vessel disease (52.4% vs 15.6%, p = 0.0001), and more frequently had lower left ventricular ejection fraction (45% vs 50%, p = 0.002). Registry patients were treated with PCI (n = 234), coronary artery bypass grafting (n = 136), or medically only (n = 226). In-hospital mortality rate among all patients who consented was 2.2% and was higher among registry than among randomized patients (4.0% vs 1.6%, relative risk 2.45, p = 0.001). Within the registry, covariate-adjusted risk of mortality was lower among patients who were treated by PCI compared with those who were treated medically (odds ratio 0.21, p=0.03). In conclusion, angiographic ineligibility for randomization in a large, prospective, primary PCI trial was associated with numerous high-risk demographic characteristics and higher short-term mortality. Nonetheless, the overall survival of patients who were referred for primary PCI was excellent, and performance of primary PCI versus medical therapy was associated with increased survival among excluded patients.