Abstract: Social insurance is often linked to marriage. I model how such linkage affects the marriage market and provide evidence exploiting Sweden’s elimination of survivors insurance. By affecting the wedge between cohabitation and marriage, this elimination altered the composition of married couples up to 50 years before expected payout, especially among couples with high husband mortality risks. Further, by reducing marital surplus, it induced divorces and intra-household redistribution towards wives. Because survivors insurance subsidized couples with highly unequal earnings, its elimination also raised the long-run assortativeness of matching. Such marriage market responses affect the optimality of linking social insurance to marriage.