Abstract Although there is evidence that the number and ages of children influence marital stability, studies have not systematically tracked the risk of marital disruption throughout the child-rearing years. This study uses marital and fertility histories from the June 1985 Current Population Survey to examine this issue. Continuous-time regression models with ages and numbers of children as time-varying covariates are estimated. Net of controls for age at marriage, year of marriage, education, and marital duration, stability increases with family size up to the third child but starts to decline as family size reaches five or more children. Aging of children is disruptive until the youngest child reaches adulthood, after which marriages become much more stable. Arrival and aging of children is an important dynamic with strong implications for marital stability.