The outcomes for young offenders (N= 37) of two teaching modules for reducing homophobia were evaluated using a multigroup pretest-posttest design. Dependent variables were cognitive, affective, and behavioral self-report measures, as well as short-story responses. Cognition was assessed by the Modified Attitudes Toward Homosexuality Scale (Price, 1982). Affects of homophobic guilt, homophobic anger, and delight were measured by the Affective Reactions to Homosexuality Scale (Van de Ven, Bornholt, & Bailey, in press). Behavioral intentions were assessed by the Homophobic Behavior of Students Scale (Van de Ven et al., in press). Interventions took two forms: a New South Wales Department of School Education module and a Community Care Schools module. The latter, which specifically addressed maintenance factors of juvenile offender homophobia, was anticipated to result in better outcomes. ANCOVAs and a difference of proportions test revealed that the Community Care Schools module was superior in terms of producing less commitment to homophobic behavior (p < .005), more positive written responses (p < .001), and more positive delight scores (p < .05). Implications for further interventions and research are discussed.