Although Bandura conceptualized moral disengagement as a global dimension encompassing a set of mechanisms, the validated factorial structures of the existing scales have not fully matched the conceptual model so far. The present study attempts to narrow the gap between the moral disengagement conceptual model and its measurement tools by testing five factorial structures based both on theory and previous empirical work. A sample of 1107 pupils from Grades 4 to 6 completed the French version of the Moral Disengagement in Bullying Scale (Fr-MDBS). Confirmatory factorial analyses revealed that a second-order factor structure, congruent with the conceptual model, was better than a single-factor and as good as a first-order multi-factorial structure. Results also supported both construct validity (gender comparison) and predictive validity of the scale. Moral disengagement mechanisms were associated with self- and peer-reported bullying and victimization, defender, and outsider behaviors. Euphemistic labeling, diffusion of responsibility, and distorting consequences were also related to bullying whereas moral justification and victim attribution were related to both bullying and victimization. Euphemistic labeling, diffusion of responsibility, distorting consequences, and victim attribution were negatively associated with defender behavior and positively related to outsider behavior. Advantageous comparison was not related to any of the examined variables. The Fr-MDBS seems to be a useful assessment tool to validly examine either the mechanisms or the global dimension of moral disengagement.