The emergence of radical wing candidates has become a global phenomenon. Previous studies on the radical candidates or parties focus on their life cycle, supporters, or the interaction with mainstream parties. How does the entrance of the radical candidate influence the election result? Contrast to conventional wisdom that the extreme splits the votes, we argue that the entrance of the extreme candidate makes perception on the moderate candidate on the same side as more moderate. Based on the extremeness aversion theory in consumer psychology, we hypothesize that the entrance of extreme candidate reframes the endpoint of ideological spectrum among available candidates, which makes the moderate one on the same side perceived as an intermediate option, and therefore enhances his or her attractiveness especially among the moderate voters. Through survey experiment concerning the unification-independence issue dominating in Taiwans elections, we provide empirical support for the extreme-ness aversion effect. This result challenges the assumption widely accepted in spatial models that the perceived policy positions of each candidate is independent to each other. The theory also predicts how the mainstream parties may strategically foster or repress their radical counterparts. When the number of moderate voters in a district reaches certain level, the mainstream party can manipulate the election result by allowing the entry of radical wing candidates and even without really changing their own manifesto.