Objectives: Perioperative bleeding is common among patients undergoing cardiac surgery; however, the definition of perioperative bleeding is variable and lacks standardization. We propose a universal definition for perioperative bleeding (UDPB) in adult cardiac surgery in an attempt to precisely describe and quantify bleeding and to facilitate future investigation into this difficult clinical problem. Methods: The multidisciplinary International Initiative on Haemostasis Management in Cardiac Surgery identified a common definition of perioperative bleeding as an unmet need. The functionality and usefulness of the UDPB for clinical research was then tested using a large single-center, nonselected, cardiac surgical database. Results: A multistaged definition for perioperative bleeding was created based on easily measured clinical end points, including total blood loss from chest tubes within 12 hours, allogeneic blood products transfused, surgical reexploration including cardiac tamponade, delayed sternal closure, and the need for salvage treatment. Depending on these components, bleeding is graded as insignificant, mild, moderate, severe, or massive. When applied to an established cardiac surgery dataset, the UDPB provided insight into the incidence and outcome of bleeding after cardiac surgery. Conclusions: The proposed UDPB in adult cardiac surgery provides a precise classification of bleeding that is useful in everyday practice as well as in clinical research. Once fully validated, the UDPB may be useful as an institutional quality measure and serve as an important end point in future cardiac surgical research. Copyright © 2014 by The American Association for Thoracic Surgery.