Background: Persons with chronic pain often report problems with cognitive abilities, such as memory or attention. There is limited understanding of whether objective performance is consistent with subjective reports, and how psychological factors contribute. We aimed to investigate these relationships in a group of patients expressing cognitive concerns, and evaluate the utility of self-report tools for pain management settings. Method: Participants with chronic pain (n = 41) completed standardized neuropsychological tests, and self-report measures of cognitive functioning, pain, mood and sleep, as part of a broader study investigating cognitive performance in pain. Results: Average neuropsychological test performance was subtly below normative means (within one standard deviation). Twenty-five percent of the sample scored substantially below age-adjusted norms on one or more objective tests. There were moderate-to-large associations between objective performance (e.g. Trail-Making B) and subjective cognitive complaints (e.g. Everyday Memory Questionnaire – Revised), controlling for age and education level. This was moderated by anxiety, such that subjective–objective relationships were particularly strong in those with higher anxiety. Poorer test performance was associated with higher pain intensity and catastrophizing. Subjective–objective cognition relationships remained after controlling for catastrophizing. Conclusion: Patients’ self-reported cognitive concerns concurred with objectively measured performance, independent of age, education and catastrophizing. Moreover, those with severe anxiety were more accurate in predicting their cognitive performance. The findings highlight some interesting cognition–mood relationships, and suggest that easy-to-administer questionnaires, such as the Everyday Memory Questionnaire – Revised and the Behavior Rating Inventory of Executive Function – Adult Version, may be useful to capture cognitive concerns in clinical settings. Significance: Cognitive concerns in chronic pain reflected objective neurocognitive performance. This was moderated by anxiety, such that self-reported cognition was more consistent with objective performance in those with high anxiety. Our findings suggest that reported cognitive concerns should be heeded, and self-report measures may be used clinically to facilitate dialogue about cognitive functioning.