We study how relationship stability of couples is affected by an increase in fathers’ involvement in staying home from work with young children. We make use of a parental leave reform in Sweden that earmarked one month of paid leave to each parent in a regression discontinuity difference-in-differences (RD-DD) framework. Couples who were affected by the reform increased the take-up of fathers’ leave but also increased their probability of separation compared to unaffected couples. We argue that the separation effect can be explained by the degree of restrictiveness of the policy in combination with role conflicts in traditional family constellations. (JEL D13, J12, J13, J16, J32)