This study is meant to investigate the effect of source which is relevant to the target product category when people are engaged in correction. Different from Kang and Herr’s finding in 2006, we argue that correction will occur whenever any bias is identified and removed out of the source even if the source is relevant to the target product category. That is, based on our study, relevance, which is the factor Kang and Herr consider to determine the occurrence of correction, is not sufficient to predict the source effect in the relevant context. However, it is the Flexible Correction Model that can provide the reasoning and explanation to the correction process. Source-salient persuasive message is adopted to examine our hypotheses. We chose two endorsers evaluated the same relevance to the target product category but different favorability, which serves as an irrelevant bias to effect product judgment. A part of participants were manipulated to be aware of not only the relevant but also the irrelevant characteristic of the endorser. According to the FCM, the irrelevant factor of the source served as a bias and triggered correction to occur when people are in high involvement to correct. The finding that people will also correct the source relevant to the target product category could provide marketers some insight on product promotion when the source relevant to target product category is adopted.