Measuring “outcomes” is at the heart of this Society’s mission and of efforts to improve health and health care delivery.  Despite this central role, there is no common agreed-upon definition as to what is meant by outcomes.  For example, for some commentators, outcomes refer uniquely to quality-of-life and survival of individual patients; this thinking underlies the US Patient Centered Outcomes Research Institute.   For others, including those doing economic evaluation, outcomes may refer to the average health benefit groups of patients. Yet others use “outcomes” to refer to aspects of functioning of the health care system.   This lack of consistency does little to illuminate the challenges in equitably delivering timely, high quality, and affordable health care.  In this presentation, the authors present a typology of outcomes for health research along with and relevant examples.   At the most granular level, endpoints in randomized trials are often clinical outcomes which are characterized as immediately observable  - “hard” - such as hospitalization, death or functional status, or latent  - “soft” - such as quality-of-life, pain, or satisfaction.  At the next level are health outcomes which are the results of care delivered in actual practice and can be subdivided into: treatment outcomes which reflect the intended and unintended medical consequences of undergoing therapy and patient outcomes which reflect the impact on patients of undergoing care in the real world.  System outcomes can be thought of as the impact of delivering care to a group of patients and are measures of the degree of functionality of the health care system.  At the highest level are societal outcomes, which measure the impact of health on the wellbeing of society.  Consensus as to what is meant by “outcomes” would be an important step towards improving the quality of the discourse and critical thinking in this area.