This study explores the news media's ability to activate racial attitudes via stereotypic portrayals of minorities in common local crime coverage. The central hypothesis is that crime news containing minority suspects primes racial attitudes, which are subsequently brought to bear on evaluations of political candidates. In an experiment, subjects were shown no crime story, a story with nonminority suspects, or a story featuring minority suspects. President Clinton's support suffered when any crime story was present, but his support was lowest among those who saw news with minority suspects. Evaluation of Clinton's performance on crime was primed powerfully by exposure to crime news, and this effect was largest when the suspects in the story were nonwhite. Spreading activation to performance on welfare, another ''race-coded'' issue, was also evident among those exposed to racially stereotypic crime stories. Finally, among whites, exposure to minority suspects boosted the importance of the president's concern for whites as a predictor of his overall support. These results suggest that implicitly racial issues are connected in memory and can be simultaneously activated by common news coverage. The findings prompt further consideration of the political impact of stereotype-reinforcing news.The influence of the news media is of continuing interest to those who would hope to understand the dynamics of political behavior. Empirical evidence provides ample support for Cohen's (1963) claim that the news might not tell us what to think, but it does help determine what we think about. In particular, we know that coverage of an issue boosts the impact of a candidate's performance on that issue as a criterion for her overall nicholas a. valentino is an assistant professor of Communication Studies and fac-