The health field is being subjected to a dictate that policy, practice and research should be informed by evidence. The mere generation of evidence, however, does not mean that policy and practice will act upon it. Utilisation and application of research findings (often equalled with 'evidence') is a political process following rationalities that are not necessarily similar to those of researchers. In response to this issue that evidence does not naturally finds its way into policy and practice (and back into research), the concept of 'knowledge translation' is becoming increasingly popular. In this article we demonstrate that 'translation' can have different meanings, and that current perspectives (both Knowledge Translation and the Actor-Network Theory) do not reflect appropriately on actions that can be taken at the nexus between research, policy and practice in order to facilitate more integration. We have developed seven conceptual categories suggesting different action modalities. Actors and actants in this game should be aware of the complex political nature of these modalities.