RECENT DATA1-4 suggested that the current health care system is providing inadequate care for dying patients, and numerous organizations5-8 have called for improvements in end-of-life care. It was proposed that external quality measures could be used to improve care at the end of life9-11 by holding health care organizations accountable for quality medical care for dying patients and by allowing consumers to use the results of these measures to choose among different health care plans. Others6,9,12 focused efforts on identifying domains relevant to this patient population (see Table 1 for examples) and on developing and identifying instruments to assess specific outcomes within these domains for quality improvement efforts.6,10,13-15 In this article, we argue that end-of-life care is fundamentally different from other medical care that is measurable by existing quality indicators. To expect measurable improvements across multiple domains in a process as complex and variable as end-of-life care either by external quality measures or through internal quality improvement efforts is optimistic at present. Many of the domains identified as candidates for measurement in end-of-life care have never been shown to be improvable by interventions available to health care professionals and organizations. Efforts to improve the quality of end-of-life care should be focused specifically on identifying and measuring those outcomes that are known to be amenable to change and on improving processes of care that are known to affect these outcomes.