Objectives: The literature on structural brain health in people with rheumatoid arthritis (RA) is very limited. Our aim was to investigate whether there are any brain phenotype differences between people that are positive rheumatoid factor (RF+) and/or people that self-reported RA, compared with people who are not RF+ and do not report RA (‘healthy controls’). Methods: We tested for cross-sectional associations between RA and/or RF+ vs. several MRI brain volume variables in the UK Biobank (n=37, 379 participants, out of which 432 reported RA and 1,833 were RF+). We adjusted for age, sex, and assessment centre in the partially adjusted models, while in the fully adjusted models we additionally controlled for deprivation, smoking and cardiometabolic conditions. Results: Those that self-reported RA were more likely to have increased log white matter hyperintensity volumes (WMH standardised β= 0.08, 95% CI= 0.021; 0.14, p<0.01), and smaller volumes of the amygdala (β=-0.09, 95% CI= -0.16; -0.02, p<0.01) in the partially adjusted analyses. In the fully adjusted model RA remained significantly associated with amygdala volume (β=-0.09, 95% CI= -0.16; -0.02, p<0.01). No significant associations with RF+ status were found. Conclusion: Our study provides evidence for differences in structural brain volumes that are known to be related to psychological and cognitive health in those with vs. without RA. Since previous literature suggests a potential role for the amygdala in pain modulation and mental health, and an increased risk for stroke associated with WMH, we believe that our findings may be of clinical importance to people with RA. Further research is necessary to confirm these associations and asses their functional implications.