OBJECTIVE: To examine how self-efficacy for arthritis pain relates to the perception of controlled laboratory pain stimuli. METHODS: Forty patients with osteoarthritis completed self-report measures of self-efficacy for arthritis pain. They then participated in a single experimental session in which measures of thermal pain threshold and tolerance were collected, as well as measures of the perceived intensity and unpleasantness of a range of thermal pain stimuli. RESULTS: Correlational analyses revealed that patients reporting high self-efficacy for arthritis pain rated the thermal pain stimuli as less unpleasant than those reporting low self-efficacy. When subjects scoring very high and very low in self-efficacy were compared, it was found that subjects scoring high on self-efficacy for arthritis pain had significantly higher pain thresholds and pain tolerance than those scoring low on self-efficacy. CONCLUSIONS: These results indicate that self-efficacy for arthritis pain is related to judgments of thermal pain stimuli. Implications for the understanding of arthritis pain and for future laboratory research are discussed.