ImportanceChild physical and emotional abuse and neglect may affect epigenetic signatures of accelerated aging several years after the exposure.ObjectiveTo examine the longitudinal outcomes of early-childhood and midchildhood exposures to maltreatment on later childhood and adolescent profiles of epigenetic accelerated aging.Design, Setting, and ParticipantsThis cohort study used data from the Future of Families and Child Wellbeing Study (enrolled 1998-2000), a US birth cohort study with available DNA methylation (DNAm) data at ages 9 and 15 years (assayed between 2017 and 2020) and phenotypic data at birth (wave 1), and ages 3 (wave 3), 5 (wave 4), 9 (wave 5), and 15 (wave 6) years. Data were analyzed between June 18 and December 10, 2023.ExposuresEmotional aggression, physical assault, emotional neglect, and physical neglect via the Parent-Child Conflict Tactics Scale at ages 3 and 5 years.Main Outcomes and MeasuresEpigenetic accelerated aging (DNAmAA) was measured using 3 machine learning–derived surrogates of aging (GrimAge, PhenoAge, and DunedinPACE) and 2 machine learning–derived surrogates of age (Horvath and PedBE), residualized for age in months.ResultsA total of 1971 children (992 [50.3%] male) representative of births in large US cities between 1998 and 2000 were included. Physical assault at age 3 years was positively associated with DNAmAA for PhenoAge (β = 0.073; 95% CI, 0.019-0.127), and emotional aggression at age 3 years was negatively associated with PhenoAge DNAmAA (β = −0.107; 95% CI, −0.162 to −0.052). Emotional neglect at age 5 years was positively associated with PhenoAge DNAmAA (β = 0.051; 95% CI, 0.006-0.097). Cumulative exposure to physical assault between ages 3 and 5 years was positively associated with PhenoAge DNAmAA (β = 0.063; 95% CI, 0.003-0.123); emotional aggression was negatively associated with PhenoAge DNAmAA (β = −0.104; 95% CI, −0.165 to −0.043). The association of these measures with age 15 years PhenoAge DNAmAA was almost fully mediated by age 9 years PhenoAge DNAm age acceleration. Similar patterns were found for GrimAge, DunedinPACE, and PhenoAge, but only those for PhenoAge remained after adjustments for multiple comparisons.Conclusions and RelevanceIn this cohort study, altered patterns of DNAmAA were sensitive to the type and timing of child maltreatment exposure and appeared to be associated with more proximate biological embedding of stress.