Objective: The aim of this retrospective cohort study was to examine whether adherence to metformin treatment may be associated with lower onset of rheumatoid arthritis (RA).Method: Using the computerized databases of a 2.3-million state-mandated health services organization in Israel, we identified incident RA cases among a cohort of 113 749 adult patients who initiated metformin therapy between 1998 and 2014. Adherence was assessed by calculating the mean proportion of follow-up days covered (PDC) with metformin.Results: During the 18 year study period, there were 558 incident RA cases (61 per 100 000 person-years). Adherence to metformin treatment was associated with a lower risk of developing RA, with the lowest risk recorded among patients with a PDC of 40-59% [adjusted hazard ratio (HR) 0.62, 95% confidence interval (CI) 0.45-0.84] compared with non-adherent patients (PDC < 20%). A mean daily metformin dose of 2550 mg or more was also associated with a lower risk of developing RA (adjusted HR 0.62, 95% CI 0.46-0.84) compared to a daily dose of 850 mg or less. In stratified analyses by gender, the negative association between adherence and the risk of RA was limited to women alone.Conclusions: Adherence to metformin treatment is associated with a reduced risk of developing RA in women. Further studies are needed to assess the effect of metformin on RA development in other patient populations.