International migrants are increasingly shaping the population health of high-income countries, yet existing research has treated immigrants as a single, homogenous group when estimating their impact on national life expectancy. Moreover, European research has only studied the impact of immigrants up until 2019—before the COVID-19 pandemic. Using Swedish total population administrative register data from 2010 to 2023, we calculate period life expectancy at age 1 for immigrant subpopulations defined by permit status (refugee; other permit) and for permit status by region of birth (Nordic, Other European, and non-European), applying the Arriaga decomposition to estimate each subgroup's contribution to national life expectancy. From 2010 to 2019, refugees were the primary driver of the positive impact of immigrants on Sweden’s national life expectancy, exhibiting consistently higher life expectancy than both non-migrants and immigrants with other permits. In 2020, COVID-19 struck refugees disproportionately hard. They exhibited life expectancy losses up to four times higher than non-migrants and other permit migrants, and their positive effect on life expectancy was temporarily reversed. The reversal was most pronounced among non-European refugees. Life expectancy recovery among refugees was achieved by 2023 but lagged 2-years behind non-migrants and other permit migrants. Demographic and public health research should routinely disaggregate immigrant populations accordingly. COVID-19 showed that the positive contribution of refugees is fragile. Targeted policies are required to protect refugee populations, particularly during periods of national crisis, if the positive and growing contribution of immigrants to Sweden’s life expectancy is going to be sustained.