The present study examined whether positive biased retelling changes one’s implicit valence. Participants completed the Go/No-go Association Task (GNAT) to measure their implicit valence toward entrance examinations, before and after story retelling session. In the session, the participants in the biased retelling condition (n＝37) talked about their experiences preparing for university entrance examination as “happy” experiences, and the participants in the unbiased retelling condition (n＝36) talked about their actual experiences. After the session, implicit valence became positive in the biased retelling group, whereas it did not change in the unbiased retelling group. The result shows that implicit valence may be changed by biased retelling.