This study investigated the family-related origins of reactive and proactive aggression in adolescents. No study thus far has examined 1) whether parenting and attachment have a joint effect on reactive and proactive aggression; 2) how marital relationships may be related to these aggression subtypes; 3) how different family subsystems, such as the parent-child and marital subsystems, affect one another in their relationships with adolescents' reactive and proactive aggression; 4) whether family variables relate equally to reactive and proactive aggression; and 5) whether the relationships between these family variables and aggression subtypes vary between boys and girls. This study used a joint residual regression method to address these research gaps and examine how family variables, including parenting styles, attachment, interparental conflict and appraisals to inter-parental conflict such as self-blame and perceived threat, influence reactive and proactive aggression in adolescence. Participants were 1,865 adolescents (aged 10-16 years) and their fathers and mothers recruited from three junior middle schools in Guangzhou, China. Adolescents reported their own reactive and proactive aggression, inter-parental conflict, their appraisals to parental conflict (e.g., self-blame and perceived threat), perceived parenting styles, and attachment with fathers and mothers. Fathers and mothers reported reactive and proactive aggression of adolescents and the parenting styles of themselves and their spouses. Existing findings suggested that parenting styles and attachment may influence children's outcomes independently, and that parenting styles are moderated or mediated by attachment. These three possible models were examined in terms of their relationships with reactive and proactive aggression after taking inter-parental conflict and appraisals into account. Support was found for the independent predictor and moderation models, and the results varied on genders of the adolescent, parenting styles and aggression subtypes. Further, attachment was found to have a moderation effect on the relationships between authoritarian parenting style and reactive aggression in girls, between authoritative parenting style and reactive aggression in boys, and between permissive parenting style and proactive aggression in boys. Secure attachment's buffering effect on reactive aggression was only evident when authoritarian parenting was low for girls and when authoritative parenting style was high for boys. Insecure attachment was only positively related to boys' proactive aggression when permissive parenting style was high rather than low. In the independent models, authoritarian and permissive parenting styles were positively related to both reactive and proactive aggression. Authoritative parenting style was not significantly related to either reactive or proactive aggression after other family variables were considered. Attachment was mainly negatively related to selfreported reactive aggression but not proactive aggression, except in models with permissive parenting styles. Based on the relationships between attachment and parenting styles, a mediation model of inter-parental conflict through parenting styles, attachment and appraisals such as self-blame and perceived threat was developed separately for girls and boys. Parenting styles, attachment and appraisals completely mediated the positive effects of interparental conflict on reactive and proactive aggression. Moreover, self-blame had only a positive effect and mediated the effect of inter-parental conflict on self-reported reactive and proactive aggression in girls but not in boys. Perceived threat had a positive effect and mediated the effect of inter-parental conflict on reactive but not proactive aggression. In addition, the family variables tended to relate more strongly to reactive than proactive aggression, especially when adolescents reported the aggression subtypes. These results suggest that the patterns of how family variables mutually influence their effects on aggressive behaviors in adolescents vary according to the aggression subtypes and gender of the adolescent. Only when both marital and parent-child relationships are considered can the family-related influences on reactive and proactive aggression in adolescents be fully captured.