One explanation for successful plant invaders is that they evolved to be more competitive. An intuitive prediction of this Evolution of Increased Competitive Ability (EICA) hypothesis never previously tested is that invasive populations should outcompete their native ‘ancestors’ in a common environment. We tested this idea in a diallel competition experiment with Alliaria petiolata where offspring from native and invasive populations were grown alone or in all pairwise combinations. While without competition, there were no differences between native and invasive populations, native populations outperformed invasive ones when competing against each other. Our results contradict the EICA hypothesis and we conclude that it does not not hold for Alliaria petiolata. Instead, we formulate a new ERCA (Evolutionary Reduced Competitive Ability) hypothesis: if there is less competition in the invasive range and competitive ability involves traits that have a fitness cost, then selection might act against it, thereby reducing intraspecific interactions too.