Differences in morphological or ecological traits expressed by exotic species between their native and non-native ranges are often interpreted as evidence for adaptation to new conditions in the non-native ranges. In turn this adaptation is often hypothesized to contribute to the successful invasion of these species. There is good evidence for rapid evolution by many exotic invasives, but the extent to which these evolutionary changes actually drive invasiveness is unclear. One approach to resolving the relationship between adaptive responses and successful invasion is to compare traits between populations from the native and non-native ranges for both exotic invaders and congeners that are exotic but not invasive. We compared a suite of morphological traits that are commonly tested in the literature in the context of invasion for three very closely related species of Centaurea, all of which are sympatric in the same native and non-native ranges in Europe and North America. Of these, C. solstitialis is highly invasive whereas C. calcitrapa and C. sulphurea are not. For all three species, plants from non-native populations showed similar shifts in key traits that have been identified in other studies as important putative adaptive responses to post-introduction invasion. For example, for all three species plants from populations in nonnative ranges were (i) larger and (ii) produced seeds that germinated at higher rates. In fact, the non-invasive C. calcitrapa showed the strongest trait shift between ranges. Centaurea solstitialis was the only species for which plants from the non-native range increased allocation to defensive spines, and allocated proportionally less resources to reproduction, patterns contrary to what would be predicted by theory and other empirical studies to enhance invasion. Our results suggest caution when interpreting the commonly observed increase in size and reproductive capacity as factors that cause exotics to become invaders.