Intensive care units provide focused, aggressive medical intervention to critically ill patients. Physicians responsible for ICU triage must decide which patients are sick enough to require this level of care and which can be managed on the general wards. While some patients are too well for the ICU, intensivists increasingly rely on another category, “too sick to benefit,” when denying ICU admission, even if beds are readily available. Recent studies indicate that between 19 and 37 percent of patients refused ICU (...) admission were declined because they were thought too sick to benefit from it, suggesting that physician use of this category is common in ICU triage.The idea of being too sick to benefit may seem paradoxical given that ICUs exist to treat the sickest of the sick. There is, however, increasing awareness that some diseases progress despite maximal intervention. Although there have been systematic attempts to define these diseases—most notably during the medical futility debates of the 1980s and early 1990s—there is little evidence about which conditions make a patient too sick to benefit from ICU admission. In the absence of a clear understanding of which diseases progress despite maximal care, ICU triage under the category “too sick to benefit” is currently done on a case‐by‐case basis. Contemporary decisions about who is too sick to benefit thus raise a number of ethical issues about what constitutes standard of care, the role of health care providers' judgments of quality of life in triage, and the just allocation of resources.Addressing these ethical concerns requires us to better define the population of critically ill adults who are too sick to benefit—a conceptual and empirical project. In this article, I recommend employing a diagnostic concept from the neonatal literature: namely, a lethal disease. (shrink)