Objectives: To adapt the Back Beliefs Questionnaire (BBQ) to measure illness beliefs in people with osteoarthritis of the hip or knee, and to measure the psychometric properties of this new 'Hip and Knee Beliefs Questionnaire' (HKBQ) together with the 'Exercise Attitude Questionnaire-18' (EAQ-18), which measures exercise-related treatment beliefs. Method: Cognitive debriefing through semi-structured interviews was used to assess the face validity of the HKBQ and the EAQ-18. Both questionnaires were tested on people with hip or knee osteoarthritis enrolled into a phase II randomized controlled trial examining the effectiveness of an evidence-based booklet in changing illness and treatment beliefs. Items in these outcome measures were accepted if they fulfilled a series of psychometric criteria including response frequency, item-total correlation, and factor analysis. The final questionnaires were tested for reliability, responsiveness, criterion, and construct validity. Results: No items were discarded from either of the final questionnaires, although one item in the EAQ-18 was borderline. Both questionnaires showed evidence of internal consistency, test-retest reliability, face validity, and criterion validity; there was less evidence for construct validity. The responsiveness of the HKBQ showed a medium-sized change; the EAQ-18 a small change. Conclusion: Both these outcome measures can be used in future trials and other studies which measure change in illness and exercise beliefs in people with osteoarthritis of the hip or the knee. Further testing of validity and responsiveness using interventions with larger effect sizes should be performed. Trial registration: Current Controlled Trials ISRCTN24554946.