This paper develops theoretical criteria and econometric methods to rank policy interventions in terms of welfare when individuals are loss-averse. Our new criterion for “loss aversion-sensitive dominance” defines a weak partial ordering of the distributions of policy-induced gains and losses. It applies to the class of welfare functions which model individual preferences with non-decreasing and loss-averse attitudes towards changes in outcomes. We also develop new semiparametric statistical methods to test loss aversion-sensitive dominance in practice, using nonparametric plug-in estimates; these allow inference to be conducted through a special resampling procedure. Since point-identification of the distribution of policy-induced gains and losses may require strong assumptions, we extend our comparison criteria, test statistics, and resampling procedures to the partially-identified case. We illustrate our methods with a simple empirical application to the welfare comparison of alternative income support programs in the US.