In a placebo (PBO)–controlled, 24-week phase 3b trial in patients with type 2 diabetes mellitus inadequately controlled with metformin (MET) + sulfonylurea (SU) (NCT01128153), adding saxagliptin (SAXA) 5 mg/d to MET+SU significantly reduced A1C vs PBO. A post hoc analysis was performed with data stratified by patient age (<65 y, ≥65 y), race (white vs Asian), baseline A1C (<8%, 8%–<9%, ≥9%), and BMI (<30 kg/m2, ≥30 kg/m2). Across categories of age (interaction P value=0.40), race (P=0.36), baseline A1C (P=0.12), and BMI (P=0.99), A1C was reduced more with SAXA vs PBO (Tables 1, 2). Adverse events were comparable across treatment groups and categories and were reported by 58%–85% of patients. Symptomatic confirmed hypoglycaemia (fingerstick glucose ≤50 mg/dL) was reported by 2 Asian patients receiving SAXA, with baseline A1C <8% and BMI <30 kg/m2. When added to MET+SU, SAXA improves A1C across categories of age, race, baseline A1C, and BMI and is generally well tolerated.