ObjectiveTo search out whether or not years of education is causally related to rheumatoid arthritis (RA).MethodWe conducted a two-sample Mendelian randomisation (MR) analysis employing inverse-variance weighted (IVW), weighted median and MR-Egger regression analysis. We chose statistic data of years of education from the UK Biobank genome-wide association studies (GWASs) (n=293 723) as the exposure and a meta-analysis of GWASs of RA with autoantibody (n=5539) and European controls (n=20 169) as the outcome.ResultsWe selected a total of 49 single nucleotide polymorphisms as instrumental variables (IVs). The IVW method instructed an inverse causative relationship between years of education and RA (β=− 0.039, SE=0.283, p=0.008). MR-Egger regression test showed that directional pleiotropy seems not to bias the MR results (intercept=0.028; p=0.358). MR-Egger analysis demonstrated no causative relationship between RA and years of education (β=− 2.320, SE=1.709, p=0.181). However, the weighted median approach indicated a causative association between RA and years of education (β=−0.950, SE=0.355, p=0.008).ConclusionsThe MR analysis supported a potential inverse causative relationship between years of education and development of RA.