Objective: The DAS28, used to assess disease activity in rheumatoid arthritis (RA), is a composite score comprising clinical, biochemical and patient self-report measures. We hypothesized that psychological factors (cognitions and mood) would be more strongly associated with patient reported components of DAS28 than clinical or biochemical components. Methods: A cross-sectional, observational study of 322 RA patients awaiting biologic therapy with active disease (mean DAS28 = 6.0) was undertaken. Patients' illness beliefs, treatment beliefs and mood were measured using the Brief Illness Perception Questionnaire (Brief IPQ), the Beliefs about Medicines Questionnaire (BMQ) and the Hospital Anxiety and Depression Scale, respectively. Relationships between psychological factors and (i) total DAS28 (ii) individual components of DAS28 were analysed using linear regression. Results: Total DAS28 produced significant but weak associations with two of the Brief IPQ items but no associations with BMQ or HADS scores. There were larger significant associations between the patient reported visual analogue scale (VAS) with five items of the Brief IPQ and with HADS depression. Low illness coherence was associated with higher tender joint count. Three Brief IPQ items and HADS anxiety scores were significantly associated with C-reactive protein or erythrocyte sedimentation rate. No psychological factors were associated with the swollen-joint count. Conclusion: One of the subjective components of the DAS28, patient VAS, was highly correlated with cognitive factors and depression in those with severe RA. By reporting individual DAS28 components, clinicians may be better able to assess the impact of therapies on each component, adjusting approaches according to patients' needs. © 2013 American College of Rheumatology.