We used a randomized quasi-experimental design to test the effectiveness of three types of perspective-taking condition in a forgiveness education program. Allport's Contact Hypothesis was used as a framework for the study design. Eighth graders (n?=?132) in an urban Midwestern city were invited to participate. We evaluated the effectiveness of perspective-taking approaches in promoting forgiveness and reducing prejudice, anger and emotional reactivity. We also explored the effects of forgiveness education across socially and culturally diverse groups. We did not find differences between the perspective-taking conditions; however, all three groups improved on both forgiveness and prejudice. We also found the pattern of outcomes was different for the African American participants than for the European American participants. Implications for research and education are discussed.