We investigate whether paying fathers to stay at home with their newborn child affects marital stability. Our empirical analysis is based on a reform in Iceland that offered one month of parental leave earmarked to fathers with a child born on or after January 2001. This reform created substantial economic incentives for fathers to be more involved in caring for their children during their first months of life, and the take-up rate in the first year was 82.4%. We apply a regression discontinuity framework to assess the effect of this reform on the probability of separation among couples and find that parents who are entitled to paternity leave are less likely to separate. The effect persists throughout the first 15 years after the child is born. Interestingly, paternity leave has the strongest impact among couples where the mother has higher, or equal, educational attainment to that of the father.