When extending a product line, how will the addition of the new product influence the purchase pattern of the existing products in the line? The behavioral literature offers two hypotheses—similarity and extremeness aversion—that make conflicting predictions about how the addition of a new extreme option should affect the choice shares of the original options in the set. In this paper, we articulate a theoretical account, prominence detraction, that invokes attribute prominence to predict when each hypothesis is satisfied. The prominence detraction hypothesis predicts that: a) the addition of a new extreme option that scores higher on a non-prominent attribute will increase the share of the more similar intermediate option, while b) the addition of a new extreme option that scores higher on a prominent attribute will decrease the share of the more similar intermediate option. We find support for the prominence detraction hypothesis in nine studies and in two meta-analyses.