Rheumatoid arthritis (RA) has been associated with endothelial dysfunction, a pathophysiological feature of atherosclerosis. Our aim was to determine whether TNF-α blockade has a beneficial effect on endothelial function in RA. We performed a systematic review with meta-analysis to evaluate the effect of anti-TNF-α agents on endothelial function in RA patients. MedLine, Cochrane CENTRAL and SCOPUS were searched up to March 2016. Inclusion criteria were: 1) randomised controlled trial (RCT), quasi-RCT, before-after cohort study; 2) including RA patients; 3) treatment with anti-TNF-α medications; 4) evaluating the change from baseline in endothelial function. The search strategy retrieved 180 records, of which 20 studies were included in the systematic review. Pooled analysis using a random-effects model demonstrated a significant improvement in endothelial function following anti-TNF-α treatment (SDM 0.987, 95%CI [0.64-1.33], p < 0.0001). Generalisation of the results of the meta-analysis may be limited due to the presence of heterogeneity (I2 = 82.65%, p < 0.001) and evidence of possible publication bias. Meta-regression showed that endothelial function measurement technique was a significant contributor to heterogeneity. In conclusion, although limited by the methodological quality of the included studies, our meta-analysis suggests that anti-TNF-α treatment may improve endothelial function in RA patients.