Background: Fatigue is a prevalent burden in RA and a key management priority for patients, but current treatment is lacking because its mechanisms are unknown. Although pain has been identified as a possible cause of fatigue, the link between these symptoms is unclear. This project aimed to examine the prognosis of fatigue and determine the role of pain in symptom outcomes. Methods: Two disease registries were used to address these aims in early – the Norfolk Arthritis Register (NOAR) – and severe RA – the British Society for Rheumatology Biologics Register for RA (BSRBR-RA). In both registries, fatigue and pain were measured along with a variety of demographic, clinical and psychosocial data. The prevalence of high symptoms was determined in both cohorts and the prognosis of fatigue delineated in NOAR using group-based trajectory modelling (GBTM). Descriptive statistics and structural equation modelling (SEM) determined the magnitude and pathways of fatigue change following commencement of anti-TNF therapies in the BSRBR-RA. Results: High fatigue was reported by 56% of 413 eligible early-RA patients. Of those, 68% reported persistent high fatigue at year one, but GBTM showed the progression of fatigue to vary over 4 years. Key sex differences in trajectories were observed. Of 6835 eligible participants commencing anti-TNF therapies, 39% reported high fatigue at baseline, but 66% improved to a low fatigue status by six-months. SEM indicated that the effect of disease activity on changes in fatigue was indirect and that improvements were largely driven by pain. The particular importance of mental health and disability were also highlighted throughout analysis. Discussion/Conclusions: Fatigue is a common and often chronic burden in RA. The symptom is amenable to treatment by anti-TNF therapies, but improvements are not driven by reductions in disease activity. Instead pain is a mediator of improvement and likely a suitable treatment target, along with mental health and disability.