In “Association of prestroke metformin use, stroke severity, and thrombolysis outcome,” Westphal et al. reviewed data from the European Thrombolysis in Ischemic Stroke Patients collaboration and compared stroke severity (NIH Stroke Scale), 3-month functional outcome (modified Rankin score), and mortality in 757 patients with type 2 diabetes who received metformin before stroke and 1,162 patients with type 2 diabetes who did not receive metformin before stroke. The authors reported that patients on metformin had less severe strokes, better functional outcome, and lower mortality and concluded that these findings suggest metformin has a protective effect in this patient population. Although he found the findings interesting, Dasheiff believed that this conclusion was premature because (1) correlation does not imply causation, (2) the statistical analysis did not fully compensate for the myriad confounders, and (3) there is no mention of other diabetic medications such as insulin use, which, in addition to the HbA1c and blood glucose may reflect disease severity. Westphal et al. agreed that causality cannot be inferred but emphasized the strengths of their study including the large sample size and their use of propensity score matching. The authors suggested their findings may precipitate additional studies on the relationship between metformin and outcome after stroke and increased use of metformin in patients with diabetes and vascular risk factors.