Background: Pain is considered a multidimensional experience and is very common in the elderly. The pain locus of control has become essential to understand how the perceptions, expectations and beliefs are related to individual behavior, attitudes, coping and adherence of the elderly with regards to the health conditions and with the proposed treatment. Studies focused on adaptation and reliability of instruments are necessary for health professionals. Objectives: The present study performed the cross-cultural adaptation of the Pain Locus of Control Scale (C form of Multidimensional Health Locus of Control) for Brazil and assessed its intra and inter-examiner reliability among a sample of 68 elderly individuals with non-oncological pain living in the community. Methods: The cross-cultural adaptation of the scale was performed using the methodology standardized by Beaton et al. (2000)*. Pearson's correlation coefficient (PCC) and the intraclass correlation coefficient (ICC) were used for the statistical analysis (p≤0.05). Results: Average age of the subjects was 69.6 ± 5.5 years; most were women, with low levels of income and education. The average pain duration was 10.2 years and the main clinical diagnosis was osteoarthritis. The reliability of the scale was adequate, with a regular to very strong correlations (PCC = 0.60 to 0.93) and a moderate to nearly perfect ICC (0.60 to 0.93), in mainly the chance locus of control and medical and healthcare professionals locus of control subscales. Conclusion: An adequate reliability and applicability was observed in our sample after adjustments and adaptations of the scale for use in elderly