The current study compared the effects of an acceptance versus distraction rationale on coping with experimentally induced pain. Eighty participants were randomly assigned to one of five conditions: Full-Acceptance, Full-Distraction, Instruction-only-Acceptance, Instruction-only-Distraction and No-Instructions. Participants completed a simple matching task and were intermittently given the choice either to receive an electric shock and continue, or to avoid the shock and terminate the task. Only the Full-Acceptance strategy (that included experiential exercises and a metaphor) had a significant effect on task tolerance as measured by an increase in the number of shocks delivered post-intervention relative to baseline. In addition, the participants in both of the acceptance conditions showed lower levels of believability in that they were more likely to continue with the task even when reporting more pain. The results support the prediction that acceptance-based interventions work by undermining the behavioural-control functions of pain-related thoughts and feelings, and call for a systematic analysis of how metaphors and exercises work in analogue research. (c) 2007 Elsevier Ltd. All rights reserved.