Abstract: Research on children's peer aggression has primarily focused on physical forms of aggression. Recently, however, the literature on children's aggression has widened its focus to include other forms of aggression relevant to both boys and girls. Relational aggression is one form or non-physical aggression that is defined as hurting someone through their relationships with others or by hurting their social standing in their peer group (Crick, 1995). The purpose of this research was to examine the social cognitive mechanisms of moral disengagement which children sometimes use to justify their aggressive behaviors. Although previous research has suggested that these cognitive mechanisms operate in the perpetuation of children's overt aggression (Bandura, Barbaranelli, Caprara, & Pastorelli, 1996; Bandura, Caprara, Barbaranelli, Pastorelli, & Regalia, 2001), it was unknown whether these same mechanisms would operate similarly within the context of relational aggression. Because a measure did not exist to assess moral disengagement in the context of overt and relational aggression, a new measure was created. The self-report Children's Moral Disengagement in Aggression Scale (CMDAS) was developed and refined through pilot-testing with 40 fourth and fifth graders at a public elementary school. Two subscales were created; one to assess moral disengagement in each form of aggression. Participants also completed the Peer Nomination Instrument (Crick, 1997). Pilot-test results suggested that the subscales of the CMDAS had fair internal consistency, higher for boys. Mixed support was found for the construct and discriminant validity of the subscales depending on gender of the participant and subscale. Significant gender differences in moral disengagement, restriction of range in moral disengagement and both forms of aggression, and a small sample size complicated efforts to develop a reliable and valid measure for both genders. The revised Children's Moral Disengagement in Aggression Scale was administered in Study 2 to a sample of 136 fourth, fifth, and sixth graders in public elementary schools. Participants were administered the same measures in Study 1, in addition to the Test of Self-Conscious Affect for Children (TOSCA-C) (Tangney et al., 1996). The subscales of the revised CMDAS were internally consistent for both genders. As before, there was mixed support for the construct and discriminant validity of the subscales, depending on the gender and subscale. There was evidence for convergent validity of both subscales. As in Study 1, there were significant gender differences in overt aggression and moral disengagement, with boys having higher levels of both compared to girls. Hierarchical multiple regression suggest moral disengagement predicted overt aggression in the total sample. No significant relationship was found between moral disengagement and relational aggression for either gender. This research adds to the recent controversy regarding whether social-cognitive mechanisms like moral disengagement operate in girls and questions their usefulness in predicting children's relational aggression.