Background: Chronic musculoskeletal pain is a major health problem accounting for approximately one quarter of GP consultations in the UK. Exercise has been shown to be beneficial for back and knee pain. However, poor adherence to exercise and physical activity may limit longterm effectiveness. The quality of methods to measure and report exercise adherence is unknown. Objectives: To highlight quality issues in measuring and reporting exercise or physical activity adherence in RCTs of interventions for chronic musculoskeletal pain. Methods: A comprehensive, sensitive search strategy was carried out on 16 bibliographic databases and evidence-based Internet resources as part of a Cochrane review. At least two independent reviewers sifted through titles and abstracts, assessed quality and extracted the data from the included studies. Due to differences in interventions, trial participants, length of follow-up and outcome measures, quantitative pooling of the study results on exercise adherence was not possible. Results: Forty-two RCTs that aim to improve exercise or physical activity adherence for people with chronic musculoskeletal pain were included in the review. We found the following problems with adherence measurement and reporting: Measurement: no consistency in measuring adherence to exercise or physical activity; most measures were self-reported, and only one study used motion sensors; attendance was often used as a measure of adherence; most trials measured adherence in the short term only, and thus the usefulness of measures to assess long-term adherence to physical activity is unknown. Reporting: there was poor quality reporting of adherence measures; most studies failed to explicitly state an aim to improve adherence; adherence data were often not reported when there were no statistically significant differences between trial groups; searching for trials of adherence is challenging, as adherence key terms may not be mentioned in the abstract. Conclusions: Measurement and reporting of exercise and physical activity adherence is not consistent in the available literature. Standardised measures of exercise adherence, and greater consistency in their use, are needed. By highlighting these issues, we hope to raise awareness of the subject of adherence and improve the standard of measurement and reporting in this field.