Pain patients and control subjects responses to ambiguous cues were compared in two separate investigations. In the first, pain patients, control subjects and physiotherapists were asked to produce a list of spontaneous associations to ambiguous cues (such as terminal and growth). To control for mood effects the experiment was repeated with three more groups: Pain patients, osteopaths and a control group. Measures of anxiety and depression were incorporated in the analysis. Results indicate that pain patients systematically produce more pain related associations than the other groups, and that this effect is independent of anxiety and depression levels. The discussion concentrates on the implications of these findings both for the theory of pain processing and for clinical interventions.