We agree with Emily Walsh (2020) that the current preferences of patients with dementia should sometimes supersede those patients’ advance directives. We also agree that consensus clinical ethics guidance does a poor job of explaining the moral value of such patients’ preferences. Furthermore, Walsh correctly notes that clinicians are often averse to treating patients with dementia over their objections, and that this aversion reflects clinical wisdom that can inform revisions to clinical ethics guidance. But Walsh’s account of the moral value (...) of the preferences of patients with dementia suffers from three major problems: (1) it does not engage the actual practices of clinical ethics; (2) it provides an inadequate account of why these patients’ preferences matter; and (3) it offers a poor explanation of clinicians’ intuitions in these cases. Her arguments engage a philosophical debate that is largely irrelevant to clinical practice and she therefore leaves pressing real-world clinical ethics questions unaddressed. After underscoring some of Walsh’s main points, we will discuss each of these shortcomings. (shrink)