Studies suggest prenatal cannabis exposure is associated with mood/behavioral problems in children. However, it is unclear if targeting modifiable domains like sleep behaviors would improve outcomes in exposed youth. Using a causal inference framework, the effect of changing sleep‐hours on changing internalizing/externalizing problems in children was examined using the Adolescent Brain Cognitive Development™ study baseline (ages 9–10; collected during 2016–2018) and year‐1 follow‐up data (N = 9825; 4663 female; 5196 white). Average treatment effects (ATE) indicated that more sleep predicted less internalizing (ATE = −.34, SE = .08, p < .001) and externalizing (ATE = −.29, SE = .07, p < .001) problems over time. However, prenatal cannabis exposure moderated the ATE on internalizing (conditional‐ATE = .91, SE = .39, p = .019), whereby participants with exposure (n = 605) did not show any effect of changing sleep‐hours on mood (B = .09, SE = .24).