Purpose The purpose of this study was to evaluate the relationship between the number of periods children were exposed to familial homelessness and childhood aggression and victimization. Methods Survey data were obtained from 4,297 fifth-grade children and their caregivers in three U.S. cities. Children and primary caregivers were surveyed longitudinally in 7th and 10th grades. Family homelessness, measured at each wave as unstable housing, was self-reported by the caregiver. Children were categorized into four mutually exclusive groups: victim only, aggressor only, victim–aggressor, and neither victim nor aggressor at each time point using validated measures. Multinomial, multilevel mixed models were used to evaluate the relationship among periods of homelessness and longitudinal victimization, aggression, and victim aggression compared to children who were nonvictims and nonaggressors. Results Results suggest that children who experienced family homelessness were more likely than domiciled children to report aggression and victim aggression but not victimization only. Multivariate analyses suggested that even brief periods of homelessness were positively associated with aggression and victim aggression (relative to neither) compared to children who were never homeless. Furthermore, childhood victimization and victim aggression significantly decreased from 5th grade to 10th grade while aggression significantly increased in 10th grade. Conclusions Children who experienced family homelessness for brief periods of time were significantly more likely to be a victim–aggressor or aggressor compared to those who were never homeless. Prevention efforts should target housing security and other important factors that may reduce children's likelihood of aggression and associated victimization.