A priority in trust research is to deepen our understanding of trust processes: how does trust develop and break down? This requires further understanding of what actions have what effects on trust in interpersonal interactions. The literature offers a range of actions that have effects on trust, but gives little explanation of why they do so, and how the actions "hang together" in their effects on trust. The question is what different classes of trust building actions there may be. Using a "relational signalling" perspective, we propose hypotheses for classes of action that trigger the attribution of mental frames (by the trustor to the trustee), and trigger the adoption of those frames by the trustor. A survey-based empirical test of trust building actions among 449 managers in 14 European countries confirms the hypotheses. © 2010 The Association for Social Economics.