The COVID-19 pandemic has increased the necessity of setting priorities in health and beyond. To set such priorities in health, we need meaningful and consistent health measures. The global burden of disease study plays an important role during the pandemic. Disease burden is largely measured through the lens of disability-adjusted life years (DALYs). Since its conception in the early 1990s, the DALY has gained traction and increased popularity. Several revisions of the value assumptions built into the DALY have followed. Moreover, discussions have been made of the different potential purposes of the DALYs, whether it be monitoring patterns and inequalities in disease burden, aggregating morbidity and mortality, or setting priorities within large health programs. However, much less attention has been on the main assumption of the DALY, that is, that DALYs = YLDs + YLLs. To elaborate, the DALY is the aggregation of years lived with disability (YLDs) and years of life lost (YLLs). However, are YLDs and YLLs entities that can be aggregated? Epicurus famously argued that existence is necessary for anything to matter to you, therefore, death is nothing to you. If Epicurus was right that death (i.e., the incident of death, not the process of dying) is not bad for those who die, then DALY’s addition of years lost by death may not give a meaningful result. There are at least three responses to this foundational DALY problem, of which a counterfactual account of harm might be the most viable approach. This counterfactual strategy, however, comes with severe challenges. I argue that Epicurus was correct in claiming that the incident of death is not bad for the dying or the dead. Death, for epicureans, means permanent non-existence, and hence, it remains questionable whether YLD + YLL makes sense.