Boys' fighting was assessed at ages six, eight, and nine. The boys (N = 69) had been selected from the 30% most disruptive children in kindergartens from low socioeconomic neighborhoods. Twenty-three percent of these disruptive boys were rated as high fighters on three assessments ("stable high fighters"), and 28% were rated as high fighters on two of the three assessments ("variable high fighters"). Forty-two percent were rated as high fighters only one out of three assessments ("occasional high fighters") and 7% were never rated as high fighters. Only high fighting in two successive years significantly increased the risk of being rated a high fighter in a following year. At age 10, stable high fighters (high fighters at ages 6, 8, 9) were perceived by teachers, peers, mothers, and the boys themselves as more disruptive and more antisocial than occasional high fighters. These results show an impressive self-other agreement in boys who have adopted a physically aggressive life style from an early age. The three groups did not differ on individual family demographic characteristics, but stable high fighters had a higher mean on an index of family socioeconomic disadvantage. Results indicate that the aggression scales which include only a few physical aggression items and many disruptive items (oppositional behavior, rejection, hyperactivity, inattention, etc.) probably aggregate two kinds of disruptive boys, the high-frequency fighters at high risk for stable disruptive, physically aggressive, and antisocial behaviors, and the disruptive low-frequency fighters who are at a lower risk of stable disruptive behavior and at a lower risk of early antisocial behavior.