In this chapter, we investigate the contribution of changes in union experience to increases in parental separation. We ask whether increases in parental separation might not have been greater if couples had not postponed childbearing or had fewer opportunities to test their relationship before becoming parents. We investigate the question in two contexts where marriage and childbearing were dramatically delayed, while at the same time young people continued to leave home and enter partnerships at relatively young ages, i.e., late twentieth century Sweden and Norway. We use union and birth histories from the Generations and Gender Surveys (conducted in 2012–2013 and 2007–2008, respectively) to estimate separation risks for parents who had their first child between 1960 and 10 years before the interview. We decompose those risks in terms of the parents’ union experiences prior to first birth, to identify those that contributed to or suppressed increases over time in parental separation. We conclude that the increasing differentiation of parents’ union experience contributed to increases in parental separation. Pure compositional shifts did not, though they likely compensated each other, some generating an increased likelihood, others a decreased likelihood of parental separation.