This research proposes to introduce the concept of stereotype to definebrand gender and to make a new contribution on the analysis of cross-gender extension evaluation. The results of an experiment, made on two product categories and considering the two possible directions of these extensions – from men to women and from women to men – reveal that the perceived fit between the cross-gender extension and the brand is more positive when the brand gender is non-stereotyped and, surprisingly, when the brand extends from the female to the male market. The interaction effect suggests also that the impact of the cross-gender extension direction is more important in the case of a brand with a non-stereotyped gender. Those results challenge previous research. A concluding discussion lays out recommendations for business.