The relation across anemia, hemorrhagic complications, and mortality associated with percutaneous coronary intervention (PCI) is unclear. We reviewed the Randomized Evaluation in PCI Linking Angiomax to Reduced Clinical Events (REPLACE)-2 Trial, which compared bivalirudin plus provisional glycoprotein IIb/IIIa blockade with heparin plus planned glycoprotein IIb/IIIa blockade in patients undergoing urgent or elective PCI. Of the 6,010 patients randomized in REPLACE-2, 1,371 (23%) were anemic. Major bleeding was more common in anemic than in nonanemic patients (4.9% vs 2.8%, p = 0.0001). In anemic patients, treatment with bivalirudin (n = 678) resulted in a lower risk of major bleeding versus heparin plus glycoprotein IIb/IIIa blockade (n = 693, 3.5% vs 6.2%, p = 0.0221). Mortality was higher in anemic patients than in nonanemic patients at 30 days (0.9% vs 0.2%, p <0.0001), 6 months (2.6% vs 0.7%, p <0.0001), and 1 year (4.3% vs 1.5%, p <0.0001). There were no differences between anemic and nonanemic patients with regard to ischemic complications at 30 days. Although anemic patients had higher mortality rates, proportions of cardiovascular and noncardiovascular mortalities were equal in anemic and nonanemic patients. In conclusion, anemic patients undergoing PCI have an increased risk of mortality and major bleeding, but not of ischemic events, and the use of bivalirudin with provisional glycoprotein IIb/IIIa blockade decreases the risk of hemorrhagic complications compared with heparin plus planned glycoprotein IIb/IIIa blockade.