Objective: Metformin is widely used as the first-line drug for type 2 diabetes mellitus and has numerous benefits apart from lowering blood glucose. However, metformin-retained regimen is challenged by newly launching, powerful glucose-lowering antiglycemic agents. This population-based cohort study examined the association between metformin adherence and the risk of dementia and Parkinson's disease (PD). Methods: Diabetic patients with metformin-included combination antiglycemic therapy were identified from the National Health Insurance Research Database and categorized into metformin-adherent and -nonadherent groups according to the medical record of the first year prescription. Patients contraindicated with metformin, severe diabetic complications, and poor drug compliance were excluded. The study outcome was the diagnosis of dementia or PD. Results: A total of 31 384 matched pairs were included after using propensity score matching and both groups were followed up for an average of 5 years. Metformin adherence was associated with a significantly lower risk of dementia (adjusted hazard risk ratio = 0.72, P < .001) but not PD (adjusted hazard risk ratio = 0.97, P = .825). Subgroup analysis revealed that the risk of dementia was significantly reduced in metformin-adherent patients, both male and female, aged >65 or ≤ 65 years, and with or without concurrent insulin treatment. This effect was not influenced by concurrent insulin treatment, which may eliminate the bias caused by the severity of diabetes mellitus. Conclusion: Despite the launching of numerous new oral antiglycemic agents, metformin may provide further benefit on lowering risk of dementia beyond conventional glycemic control according to the real-world evidence.