Similarity between a core product and an extension product is regarded as a key success factor for brand extensions. However, a review of the literature reveals that the conditions under which high similarity is beneficial are unclear. In the first experiment, the authors test three models (the brand-attitude-transfer model, the incongruity-induced-affect model, and the “you cannot be good or bad ... [Show full abstract] at everything” principle) to predict the role of similarity. They find support only for the brand-attitude-transfer model. This suggests that high similarity is only beneficial when consumers evaluate brands favorably. Furthermore, little is known about how additional information provided by the supplier affects consumers' similarity perceptions without changing the extension product itself more than on a superficial level and without emphasizing common benefits. The authors compare the conditions of high and low core product/extension category similarity. Under the former condition, common peripheral advertising cues, similarity primes, portrayals of core products, and common peripheral design cues increase evaluations of extension products when the brand is evaluated favorably. Under the latter condition, these cues differ with respect to their effect.