Mediated intergroup contact can reduce recipients’ prejudice. However, prior studies have mainly examined the effects of carefully selected stimuli, contents, and genres. We investigated whether outgroup exposure on the level of real-life, multi-channel media diets is associated with reduced prejudice. Prior scholarship offers contradictory expectations: We tested this with survey data from South Africa where the numeric group constellation and the media system should make blatant stereotyping in traditional media less likely. While social media interactions were associated with more positive attitudes, effects in traditional media varied across the type of medium and sample. This suggests that it will be difficult to associate a person’s general media usage with their prejudice. We discuss the challenges for future tests of the real-life potential of mediated intergroup contact. Altogether, our study dampens initial hopes and underlines the importance of targeted interventions and outstanding single contents.