We assess the effect of children on household dissolution and fomation by looking at how children influence the probability of divorce or cohabitation breakup, as well as the chance of repartnering in the 20 years after they were born. Using the danish registry data from 1980 to 2013 we dispose of a panel of the entire Danish population, which allows us to use multiple births (twins) as an instrument in order to obtain causal estimates of children on the likelihood of divorce. We use both repeated instrumental variable regressions as well as hazard models to identify how marital stability evolves in each year since birth was given. Initial results indicate that children have close to no effect on marital stability in the first ten years after birth but that marriages get slightly more stable in the presence of more children as these children grow older. The results can be interpreted in a framework where children cause both stress as well as increased marital cohesion. As the stress is reduced over time, cohesion effects dominate.