Two articles in this issue of JAMA1,2 make an important contribution to the growing debate about medical futility. The authors present a strong case that futile resuscitation need not be offered to patients and their families and may be denied when they request it. I have argued elsewhere that physicians might unreasonably diminish patient autonomy by acting on poorly defined notions of futility.3 These articles make an equally important observation: by offering life-sustaining treatment that they consider futile, physicians may actually undermine patient and family autonomy. Such offers send a mixed message, implying a real choice when none exists. Furthermore, physicians have a socially sanctioned responsibility to evaluate, promote, and, when necessary, protect patients' best interests. As Tomlinson and Brody note, this role inevitably involves making value judgements as the physician weighs the potential harms and benefits of any given intervention. If such value judgments are "socially validated,"