Identification of the causes of child maltreatment perpetration is prerequisite for developing efficacious prevention initiatives to reduce its occurrence. Earlier maltreatment victimization is often suggested as an important cause of subsequent maltreatment perpetration. This study investigates a) whether maltreatment victimization causes subsequent perpetration and b) whether the timing of maltreatment victimization--childhood versus adolescence--alters this relationship. We use data from the Rochester Youth Development Study, a longitudinal study begun in Rochester, New York in 1988, based on a stratified random sample of 1000 seventh and eighth graders. At the most recent followup, 80 % were reinterviewed. Child Protective Services data were collected on substantiated incidents of maltreatment victimization from birth to age 17 and on maltreatment perpetration through average age 33, n = 816. Using propensity score models to control selection effects, we find that a history of maltreatment victimization significantly increases the odds of becoming a perpetrator of maltreatment. Although childhood-limited maltreatment does not significantly increase the odds of maltreatment perpetration, maltreatment that occurs in adolescence or that begins in childhood and persists into adolescence does. Adolescent maltreatment was found to be more serious in terms of type, chronicity, and severity than childhood-limited maltreatment, offering a possible explanation for this finding. Therefore, maltreatment victimization, especially during adolescence, is a likely cause of subsequent perpetration. Clinical services to interrupt the cycle of maltreatment are needed to protect subsequent generations from experiencing maltreatment and from experiencing the health-risking behaviors that often flow from maltreatment.