Unemployment Accounts (UA) are mandatory individual saving accounts that can be used by governments as an alternative to the Unemployment Insurance (UI) system. The goal of this paper is to study the welfare implications of a shift from the current UI system to a new UA system in the United States. The UA system works as follows. During employment, the worker is mandated to make deposits into the individual saving account. The worker is entitled to withdraw payments from this account only during unemployment. In contrast, UI is funded by a payroll tax and provides benefits for a limited duration. I build an heterogeneous agents, incomplete-markets life-cycle model, in which workers face income fluctuations and unemployment shocks. UI is modeled as a choice of a replacement rate, and a time limit of unemployment benefits. UA is modeled as a choice of a deposit rate into the account during employment and a withdrawal rate during unemployment. Qualitatively, a shift from UI to UA can lead to either a welfare gain or a welfare loss depending on the role of frictions and incentives in the model. This observation puts the paper at the nexus of the macroeconomic debate on the level of disutility from work. Quantitatively, for a plausible parameterization the shift from UI to UA leads to an average welfare gain of 0.9% of lifetime consumption.