Objectives: Although acupuncture is often advocated for patients with rheumatoid arthritis (RA), its efficacy for type 2 diabetes mellitus (T2DM), a common metabolic disease among RA cohorts, has not yet been established. This retrospective cohort study aimed to determine the association between acupuncture use and the development of T2DM among them.Methods: Data were collected from 1999 through 2008 for individuals aged 20–70 years in the nationwide insurance database of Taiwan. From them, we extracted 4,941 subjects within newly diagnosed RA and being T2DM free at baseline. A total of 2,237 patients had ever received acupuncture, and 2,704 patients without receiving acupuncture were designated as a control group. All of them were followed to the end of 2013 to identify T2DM incidence. The Cox proportional hazards regression model was utilized to obtain the adjusted hazard ratio (HR) for acupuncture use.Results: Compared with the RA subjects without use of acupuncture, the incidence of T2DM was lower for those who received acupuncture, with the incidence rates of 24.50 and 18.00 per 1,000 person-years (PYs), respectively. After adjusting for potential confounders, use of acupuncture was significantly related to the lower T2DM risk, with the adjusted HR of 0.73 [95% confidence interval (CI) 0.65–0.86]. Those who used acupuncture for more than five sessions had the greatest benefit in lowering the susceptibility to T2DM.Conclusion: Adding acupuncture into conventional treatment for RA was found to be related to lower risk of T2DM among RA patients. Further clinical and mechanistic studies are warranted.