Drawing from the contact hypothesis, the present research examines the effect of positive and negative contact on intergroup attitudes of white and black secondary school adolescents in Portugal (N = 85). Both majority and minority groups are examined in a longitudinal design. Cross-lagged regression analyses shows that the "from contact to less prejudice" path is confirmed only for majority group students, for whom positive contact was effective in intergroup bias reduction. Negative contact predicts negative intergroup attitudes for both groups, with a strongest effect for the minority group member. In comparison, emerged that for majority group students, positive contact is more influential on bias reduction. For minority group students, negative contact is more powerful in increasing negative intergroup attitudes more than the positive contact is to its reduction, and the opposite path was found too, as intergroup bias was a predictor of negative contact. These results underline the importance of analysing both positive and negative intergroup contact effects on ethnic prejudice, considering that the potential of negative contact in the increase of prejudice was largely neglected in the area of intergroup relations.