Objective: Rheumatoid arthritis (RA) is a chronic, autoimmune, inflammatory disease. Studies suggest that pro-inflammatory cytokines may be attenuated by the vagus nerve through the cholinergic anti-inflammatory pathway. We aimed to evaluate the anti-inflammatory effects of short-term transcutaneous non-invasive vagus nerve stimulation (n-VNS) applied to the cervical vagus nerve in patients with RA. Method: We conducted a single-centre, open-label, preliminary proof-of-concept study of n-VNS in two cohorts of participants with RA: one with high disease activity (n = 16) and one with low disease activity (n = 20). Disease Activity Score based on 28-joint count-C-reactive protein (DAS28-CRP), cardiac vagal tone, and pro-inflammatory cytokines were measured at baseline and after 1 and 4 days of n-VNS. Results: In the high disease activity group, n-VNS resulted in reductions in DAS28-CRP (4.1 to 3.8, p = 0.02), CRP (8.2 to 6 mg/mL, p = 0.01), and interferon-γ (29.8 to 22.5 pg/mL, p = 0.02). In the low disease activity group, there was no effect on DAS28-CRP, and n-VNS was associated with a decrease in cardiac vagal tone (p = 0.03) and a reduction in interleukin-10 (0.8 to 0.6 pg/mL, p = 0.02). Participants with high disease activity had lower baseline cardiac vagal tone than those with low disease activity (3.6 ± 2 vs 4.9 ± 3 linear vagal scale, p = 0.03). Cardiac vagal tone was negatively associated with DAS28-CRP (r = -0.37, p = 0.03). Overall, n-VNS was well tolerated. Conclusion: This study provides preliminary support for an anti-inflammatory effect of n-VNS in patients with RA. These findings warrant further investigation in larger placebo-controlled trials.