Background Fatigue is reported by up to 80% of patients with rheumatoid arthritis (RA) and is a principal determinant of poor quality of life. Existing analysis of long-term fatigue has examined only average changes on a group level, which may not account for the varied progression of fatigue commonly observed in practice. This study aimed to determine individual trajectories of fatigue and whether those with the worst prognosis can be identified at initial presentation. Previous SectionNext Section Methods or Theme Participants meeting RA classification, from a long-term observational inflammatory polyarthritis cohort study (the Norfolk Arthritis Register (NOAR)) reported fatigue (visual analogue scale, VAS) annually, for five years. Group-Based Trajectory Modelling determined the fatigue trajectories of those with clinically relevant baseline fatigue (VAS ≥ 20 mm). Baseline variables, grouped as demographic, clinical and patient-reported, were compared for between-trajectory differences using descriptive statistics (p < 0.05 considered significant). Previous SectionNext Section Results Number of trajectories differed between the sexes. ‘Improved’ participants (Figure 1) were compared to those on non-improved trajectories. For both sexes, non-improvers reported significantly poorer baseline scores for disability, pain and fatigue, and were more likely to be work disabled. Additionally, female improvers reported significantly less baseline disease activity, stomach problems, sleep and use of anti-depressants or NSAIDS, while male improvers used less general analgesia. Overall, patient-reported variables characterised group membership better than clinical variables. Previous SectionNext Section Conclusions On first presentation, RA patients with significant fatigue follow distinct trajectories in relation to this symptom. Baseline characteristics can identify those likely to follow a poor prognostic path so allowing future targeting of fatigue specific interventions.