It was hypothesized that anxiety which is relevant to the source of pain exacerbates pain, whereas anxiety which is irrelevant to the source of pain reduces the experience of pain. Female subjects were given either high or low anxiety provoking information about a cold pressor task (relevant anxiety) or high or low anxiety provoking information about a potential shock (irrelevant anxiety). Subjects were then exposed to the cold pressor. The results demonstrated that subjects who were highly anxious about the cold pressor reported experiencing the most pain. Subjects who were highly anxious about the shock reported the least pain and reported significantly less pain than subjects who were highly anxious about the cold pressor. These findings clearly demonstrate that the relationship between anxiety and pain is not always positive or unidirectional.