This 24-week, double-blind, placebo-controlled, phase III trial evaluated the efficacy and safety of linagliptin in 206 Chinese patients with inadequately controlled (glycated haemoglobin [HbA1c] 7.5%-10.0%) type 2 diabetes mellitus (T2DM) receiving insulin (basal or premixed) +/- metformin. Patients were randomized (1:1) to receive linagliptin 5 mg/d or placebo. The decrease from baseline in HbA1c (primary endpoint) was greater with linagliptin than with placebo (-0.61% vs. -0.20%, adjusted mean difference -0.40%; P = 0.0016). Linagliptin demonstrated significantly greater improvement in 2-hour postprandial glucose (-1.77 mmol/L [-31.95 mg/dL]; P < 0.001), and a numerical reduction in fasting plasma glucose (-0.34 mmol/L [-6.2 mg/dL]; P = 0.2241) versus placebo. Proportionally more patients on linagliptin achieved a HbA1c reduction of >= 0.5% versus those on placebo (odds ratio 2.293, P < 0.01). Adverse events in both groups were similar, with no new safety findings or clinically relevant changes in body weight. Among investigator-defined hypoglycaemic events (linagliptin: 17.3%; placebo: 12.7%; odds ratio 1.48, P = 0.337), none were severe. In Chinese patients with T2DM, linagliptin add-on to insulin improved glycaemic control and was well tolerated, without increased risk of hypoglycaemia or weight gain.