Guided by developmental models examining the legacy of childhood caregiving environments, we examined the longitudinal pattern of associations between harsh parenting and children's internalizing and externalizing symptoms across late childhood to late adolescence. Participants included 199 youth (48.7% female, 65.3% White, 32.2% Black, 2.5% biracial) and their mothers and fathers from a diverse range of socioeconomic backgrounds. The study utilized a multi-informant, longitudinal design including five waves of data (youths' mean ages were 9, 10, 11, 17, and 18 across waves). Harsh parenting at Age 9 predicted higher levels of (a) externalizing symptoms at Ages 11, 17, and 18 and (b) internalizing symptoms at Ages 17 and 18. Developmental sensitivity analyses revealed that the magnitude of the more distal association between early harsh parenting and later internalizing and externalizing symptoms was statistically stronger as compared to more proximal associations. Bidirectional analyses revealed that externalizing symptoms at Age 9 predicted harsh parenting at Ages 9, 10, 11, 17, and 18. Whereas links between harsh parenting and internalizing symptoms were consistent with a sleeper effects model, links between harsh parenting and externalizing symptoms provided some support for both enduring and sleeper effects models. Findings inform an understanding of youth developmental sensitivity to harsh parenting and the downstream consequences of harsh parenting. Results have important translational implications, including testing the long-term efficacy of therapeutic programs.