Objective. RA is a chronic disease with frequent psychological co-morbidities, of which depression and anxiety are two common manifestations. We aimed to identify predictive factors of psychological distress in a large prospective cohort of very early RA patients. Methods. ESPOIR (Etude et Suivi des Polyarthrites Indifférenciées Récentes) is a multicentre, longitudinal and prospective cohort study of patients with early arthritis (<6 months disease duration). The study sample comprised 641 patients with very early RA according to the 2010 ACR/European League Against Rheumatism RA criteria from the ESPOIR cohort. Psychological distress was assessed over 3 years by the five-item Mental Health Inventory questionnaire at various time points (baseline, 6, 12, 18, 24 and 36 months). Logistic regression with a generalized estimating equation model was used to analyse the association of disease variables and risk of psychological distress. Results. At baseline, 46.9% of RA patients were screened as positive for psychological distress. Over 3 years, psychological distress decreased significantly, with a prevalence of 25.8% at 36 months. The HAQ Disability Index (HAQ-DI) score was the most important factor predicting psychological distress over 3 years [odds ratio 2.10 (95% CI 1.41, 3.14)–3.59 (2.29, 5.63)]. Baseline biological and radiological variables and treatment regimens were not associated with distress. Conclusion. Psychological distress in very early RA is frequent and the HAQ-DI score is a predictor of depression and anxiety in these patients. A psychological evaluation in patients with early RA is important for further individual psychiatric diagnosis and management.