AimTo evaluate the relative safety and effectiveness of bexagliflozin as an adjunct to metformin for the treatment of type 2 diabetes mellitus.MethodsIn total, 317 participants were randomized to receive bexagliflozin or placebo plus metformin. The primary endpoint was the change in glycated haemoglobin (HbA1c) from baseline to week 24, with secondary endpoints for systolic blood pressure (SBP), fasting plasma glucose and weight loss. An open label arm enrolled participants with HbA1c >10.5% and was analysed separately.ResultsThe mean change in HbA1c was −1.09% (95% CI −1.24%, −0.94%) in the bexagliflozin arm and −0.56% (−0.71%, −0.41%) in the placebo arm, a difference of −0.53% (−0.74%, −0.32%; p < .0001). Excluding observations after rescue medication, the intergroup difference was −0.70% (−0.92, −0.48; p < .0001). The open label group change in HbA1c was −2.82% (−3.23%, −2.41%). Placebo‐adjusted changes from baseline SBP, fasting plasma glucose and body mass were −7.07 mmHg (−9.83, −4.32; p < .0001), −1.35 mmol/L (−1.83, −0.86; p < .0001) and −2.51 kg (−3.45, −1.57; p < .0001). Adverse events affected 42.4% and 47.2% of subjects in the bexagliflozin and placebo arms, respectively; fewer subjects in the bexagliflozin arm experienced serious adverse events.ConclusionsBexagliflozin produced clinically meaningful improvement in glycaemic control, estimated glomerular filtration rate and SBP when added to metformin in a population of adults with diabetes.