Although existing comparative advertising research has predominantly examined the differences between simple (non-comparative) advertising and comparative advertising, these studies have failed to present consistent results regarding comparative advertising. One reason for this is that audience responses may vary depending on differences in message credibility. Therefore, this study compared the advertising effectiveness when comparative information was presented in numerical versus verbal form to enhance message credibility. Additionally, source credibility was added as another credibility variable, and the advertising effectiveness was examined when comparative information was delivered by an advertising model (endorser) versus when it was not. The results showed that presenting comparative information in numerical form was more effective than presenting it in verbal form, regardless of whether an advertising model was used. Furthermore, when comparative information was presented numerically, there was little difference in advertising effectiveness between using and not using an advertising model. When an advertising model was used in comparative advertising, advertising effectiveness was additionally enhanced for both numerical and verbal forms. Meanwhile, the additional advertising effectiveness of the advertising model (affective attitude and behavioral attitude) was greater when comparative information was presented in verbal form than in numerical form. Considering these findings, since direct comparative advertising is practically difficult, one implication is that comparative information could be presented in verbal form accompanied by an advertising model as a viable strategy.