A widely researched panacea for reducing intergroup prejudice is the contact hypothesis. However, few longitudinal studies can shed light on the direction of causal processes: from contact to prejudice reduction (contact effects) or from prejudice to contact reduction (prejudice effects). The authors conducted a longitudinal field survey in Germany, Belgium, and England with school students. The sample comprised members of both ethnic minorities (n = 512) and ethnic majorities (n = 1,143). Path analyses yielded both lagged contact effects and prejudice effects: Contact reduced prejudice, but prejudice also reduced contact. Furthermore, contact effects were negligible for minority members. These effects were obtained for 2 indicators of prejudice: negative intergroup emotions and desire for social distance. For both majority and minority members, contact effects on negative emotions were stronger when outgroup contacts were perceived as being typical of their group. Contact effects were also mediated by intergroup anxiety. This mediating mechanism was impaired for minority members because of a weakened effect of anxiety on desire for social distance. Theoretical and practical implications of these findings are discussed.