This paper investigates separation proneness after children move out of the parental homes. It contrasts two competing theories, one postulating that empty nest has a destabilizing effect on the parental couple (due to lower parental well-being and lower barriers to separation), the other reduced role strain and increased investment into the parental union (“second honeymoon” effect). Empirically, the paper uses panel data from the Generations and Gender Programme and applies within person change-score models to see how parental separation proneness develops over time. Results indicate that the probability of separation proneness declines after children move out among couples that were separation prone. No change in separation proneness is found in couples that reported no separation proneness at wave 1.