The present study investigates whether or not reading about corrupt politicians influences peoples' subsequent judgments toward political actors' supposed corruptness. We expected this media stereotype priming effect to be dependent on pre-existing implicit stereotypes. It was hypothesized that only those participants would show a media priming effect who already have a strong automatic association between 'politicians' and 'corrupt' in memory prior to reading a further facilitative article ("politicians are corrupt"). Conversely, people who do not have a comparable biased cognitive association should not. Data from an experiment support this hypothesis: We found pre-existing implicit stereotypes to moderate the media priming effect on explicit stereotypes, but only when the newspaper article covered the "corrupt politician" media stereotype with sufficient salience. Furthermore, the experiment showed that antagonistic media primes ("politicians are honest") did not produce a media priming effect at all. Antagonistic articles were simply not able to prime corruption-related memory traces.