Do regions compete, as firms do? How does one deal with the fact that regions, unlike organizations, are entities that do not act? Does it make sense to talk about the ability of regions to generate new variety? This paper aims to address these questions from an evolutionary perspective. It is meaningful to talk about regional competitiveness when the region affects the performances of local firms to a considerable degree. This is especially true when the competitiveness of a region depends oil intangible, non-tradable assets based on a knowledge and competence base embedded in a particular institutional setting that are reproduced and modified through the actions and repeated interactions of actors. Although regions are increasingly becoming collective players actively responding to an increasing exposure to extra-regional competition, the paper explains why there are serious limits in enhancing the competitiveness of regions. By doing so, it questions the usefulness of benchmarking practices with the purpose of improving regional competitiveness: there exists no 'optimal' development model, it is difficult to copy or imitate a successful model from elsewhere, and new trajectories often emerge spontaneously and unexpectedly in space.