AbstractThe evidence for differential effects of positive and negative intergroup contact on prejudice is mixed. We propose that the closeness of the relationships respondents have with contact partners can explain inconsistencies in previous findings. In three studies (total N = 953), we tested the associations between positive intimate, negative intimate, positive superficial, and negative superficial contact and attitudes toward different outgroups (immigrants and gay people). We hypothesized that: (H1) the effect of positive contact would be maximized in intimate interactions; (H2) the effect of negative contact would be maximized in superficial interactions; (H3) positive intimate and negative superficial contact would have similar power in predicting attitudes. Results always supported H1, and supported H2 and H3 only with immigrants. Findings held also when controlling for category salience and agreeableness, but not for social dominance orientation, which instead strengthened the effect of negative superficial contact. Overall, findings clarify the positive–negative contact asymmetry.