OBJECTIVES: We evaluated patients at tertiary [both percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) capable] and primary hospitals in the EARLY-ACS trial. BACKGROUND: Early invasive management is recommended for high-risk non-ST-segment elevation acute coronary syndromes. METHODS: We evaluated outcomes in 9,204 patients presenting to: tertiary sites, primary sites with transfer to tertiary sites ("transferred") and those who remained at primary sites ("non-transfer"). RESULTS: There were 348 tertiary (n = 7,455 patients) and 89 primary hospitals [n = 1,749 patients (729 transferred; 1,020 non-transfer)]. Significant delays occurred in time from symptom onset to angiography (49 hr), PCI (53h), and CABG (178 hr) for transferred patients (P < 0.001). Non-transfer patients had less 30-day death/myocardial infarction [9.4% vs. 11.7% (tertiary); adjusted odds ratio (OR): 0.78 (0.62-0.97), P = 0.026]; transferred (14.0%) and tertiary patients were similar [adjusted OR: 1.23 (0.98-1.53), P = 0.074]. Non-transfer patients had lower 1-year mortality [4.3% vs. 6.3% (tertiary); adjusted hazard ratio (HR): 0.64 (0.47-0.87), P = 0.005]: there was no difference between transferred and tertiary patients [5.2% vs. 6.3%; adjusted HR: 0.80 (0.58-1.12), P = 0.202]. Despite similar rates of catheterization, GUSTO severe/moderate bleeding within 120 hr was less in non-transfer [3.1% vs. 6.7% (tertiary); adjusted OR: 0.47 (0.32-0.68), P < 0.001], whereas transferred (6.1%) and tertiary patients were similar [adjusted OR: 0.94 (0.68-1.30), P = 0.693]. There was no difference in non-CABG bleeding. CONCLUSIONS: Timely angiography and revascularization were often not achieved in transferred patients. Non-transferred patients presenting to primary sites had the lowest event rates and the best long-term survival.