This article examines how the parental divorce process affects youth substance use at various stages relative to the divorce. With child-fixed-effect models and a baseline period that is long before the divorce, the estimates rely on within-child changes over time. Youth are more likely to use alcohol 2-4 years before a parental divorce. After the divorce, youth have an increased risk of using alcohol and marijuana, with the effect for marijuana being 12.1 percentage points in the two years right after the divorce (p = .010). The magnitudes of the effects persist as time passes from the divorce.