The present study examined whether there are distinct groups of children with reactive versus proactive motives for their aggressive behavior. We extended previous research by using a person-based analytical approach on data from a questionnaire assessing children’s motives independently from the severity of their aggression. Two competing hypotheses were tested. The both subtypes hypothesis holds that both reactive and proactive subtypes exist, as well as a mixed subtype. The reactive only hypothesis holds that only reactive and mixed subtypes exist. Hypotheses were tested on existing data from a community sample of children displaying aggression (Study 1: n = 228, ages 10–13, 54% boys), and two clinical samples of children with aggressive behavior problems (Study 2: n = 115, ages 8–13, 100% boys; Study 3: n = 123, ages 6–8, 78% boys). Teachers reported on children’s reactive and proactive motives. We selected measures available from peers, parents, teachers, and children themselves to compare the supported subtypes on variables that previous literature suggests uniquely correlate with reactive versus proactive aggression. Confirmatory latent profile analyses revealed that the both subtypes hypothesis best fit the data of all three samples. Most children were classified as reactive (55.7–61.8% across samples), with smaller percentages classified as proactive (10.4–24.1%) and mixed (18.0–33.9%). However, these subtypes only differed in expected directions on 7 out of 34 measures. Overall, results support the existence of both reactive and proactive subtypes of aggressive children, but the distinctiveness of these subtypes in terms of social-emotional characteristics warrants further study.