Aim SRT2104 is a selective activator of SIRT1. In animal models, SRT2104 improves glucose homeostasis and increases insulin sensitivity. We evaluated the tolerability and pharmacokinetics of SRT2104, and its effects on glycaemic control, in adults with type 2 diabetes mellitus. Method Type 2 diabetics with glycosylated haemoglobin (HbA1c) ≥ 7.5% and ≤10.5%, fasting glucose ≥160 and ≤240 mg dl−1, and on stable doses of metformin were evenly randomized to placebo or SRT2104 0.25 g, 0.5 g, 1.0 g or 2.0 g, administered orally once daily for 28 days. Changes in fasting and post‐prandial glucose and insulin were analyzed. Results Safety evaluation found no major differences between groups in the frequency of adverse events. SRT2104 concentrations did not increase in a dose‐proportional fashion. Significant variability in exposure was observed. Treatment with SRT2104 did not lead to any consistent, dose‐related changes in glucose or insulin. Day 28 change from baseline (mean (SD)): fasting glucose (mmol l−1) = −1.17 (2.42), −1.11 (3.45), −0.52 (2.60), −0.97 (2.83) and −0.15 (2.38) for placebo, 0.25 g, 0.5 g, 1.0 g and 2.0 g, respectively. Day 28 change from baseline (mean (SD)): fasting insulin (mmol l−1) = 1.0 (51.66), 8.9 (95.04), −6.9 (41.45), 4.1 (57.16) and 15.2 (138.79) for placebo, 0.25 g, 0.5 g, 1.0 g and 2.0 g, respectively) Treatment with SRT2104 was associated with improvement in lipid profiles. Conclusion Treatment with SRT2104 for 28 days did not result in improved glucose or insulin control which is likely due to the observed pharmacokinetics which were not dose proportional and had large between subject variability.