Background: The risk of benign brain tumors (BBT) associated with metformin use has not received much attention. Therefore, a retrospective cohort study was designed to investigate such an association in patients with type 2 diabetes mellitus (T2DM). Methods: We used the database of Taiwan’s National Health Insurance to enroll 152,176 ever users and 16,120 never users of metformin for the follow-up of incidence of BBT and a more specific outcome of cerebral meningioma. The patients were newly diagnosed with T2DM between 1999 and 2005; and they were followed up from 1 January 2006 until 31 December 2011. Hazard ratios were estimated by Cox regression incorporated with the inverse probability of treatment weighting using propensity score. Results: During follow-up, 111 never users and 557 ever users were diagnosed with BBT. For BBT, the respective incidence rates for never users and ever users were 153.95 per 100,000 person-years and 77.61 per 100,000 person-years. While ever users were compared to never users, the hazard ratio was 0.502 (95% confidence interval: 0.409–0.615). A dose-response pattern was seen when ever users were categorized into tertiles of cumulative duration of metformin therapy (cutoffs: <27.10 months, 27.10–58.27 months and >58.27 months) with respective hazard ratios of 0.910 (0.728–1.138), 0.475 (0.375–0.602) and 0.243 (0.187–0.315). For cerebral meningioma, the overall hazard ratio was 0.506 (0.317–0.808); and the hazard ratios comparing the respective tertiles to never users were 0.895 (0.531–1.508), 0.585 (0.346–0.988) and 0.196 (0.104–0.369). Conclusions: A reduced risk of BBT and cerebral meningioma is observed in metformin users in patients with T2DM.