The importance of competition has been defined as the impact or role of competition relative to the total impact of the environment, and the intensity of competition is its absolute impact. Understanding the distinction has been proposed as key in reconciling long-running ecological debates. An index of competition importance, Cimp, has been used in a number of recent studies. Rees, Childs & Freckleton (2012) present a strong attack on the competition index Cimp (and the associated interaction index Iimp) and question the underlying rationale for the concept of competition importance. We assess their critique and challenge it in a number of areas. Rees et al. conflate criticism of the index Cimp with criticism of the concept of competition importance. Their approach to assessing the properties of Cimp (and Iimp), including the use of target plant success in the absence of neighbours (PNC) as a measure of environmental severity, is biased towards demonstrating a simple linear relationship between severity and Cimp, and supporting the argument that there is no need for separate measurements of importance. We consider the proposal by Rees et al. for the use of variance partitioning to assess importance. Although providing a measure of the relative role of competition, it is unable to assess the shape of the importance-severity relationship, something central to testing important ecological theories such as those of Grime (1979) or Bertness & Callaway (1994). We discuss alternative approaches which might address this question. Synthesis. Responding to Rees et al. has been beneficial in clarifying points of difference between our approaches, and the need to visualize the shape of the importance–severity relationship. From this response we draw three broad conclusions. We need: (i) metrics of environmental severity that are independent of the success of target species; (ii) analytical approaches that avoid the statistical problems associated with ratios whilst enabling us to assess the shape of the severity–importance relationship; (iii) new data to assess the generality of proposed relationships. Studies incorporating these elements will take forward our understanding of the role of competition in plant communities.