Abstract: This article studies relationship break‐up among married and cohabiting couples, based on the Belgian data of the Crossroads Bank of Social Security. The results are based on a sample of couples marrying for the first time or starting to cohabit for the first time between 1999 and 2001. The sample is followed over time until 2013. The purpose of this study is to gain insight in relationship break‐up of married and cohabiting couples using register data. Given the fact that cohabiting couples are underestimated in official statistics because these only use offi‐ cially registered partnerships (e.g. legal cohabitation), cohabiting couples are identified in this article on the basis of their LIPRO typology giving a more correct insight in relationship dynamics of cohabiting couples. The article looks at patterns of relationship dissolution with the aid of survival analysis and a discrete time event history analysis for three relationship trajectories: marriages formed without premarital cohabitation, marriages formed after a period of premarital cohabitation and cohabitations not (yet) converted into a marriage during the observation period. The results show that cohabiting couples not marrying during the observation period, have a much lower chance to be together after 14 years compared to the married couples. The differences be‐ tween married couples with and without a period of unmarried cohabitation are more limited. Further, we find that a break‐ups are more common among couples who start living together at an early age, start from a weaker economic background or those that do not have children during the first four years. The association between relationship break‐up and these background characteristics is similar among all three relationship trajectories studied.