Aim: To investigate the safety, tolerability, pharmacokinetics and pharmacodynamics of linagliptin in patients with type 2 diabetes mellitus (T2DM). Methods: After screening and a 14‐day washout, subjects received linagliptin 2.5, 5 or 10 mg or placebo once‐daily for 28 days in this randomized, double‐blind, parallel, placebo‐controlled within‐dose groups study. Results: Seventy‐seven patients entered the study (linagliptin: 61; placebo: 16). Four patients withdrew prematurely. There was little evidence of linagliptin accumulation. Exposure, maximum and trough plasma concentrations of linagliptin increased less than dose‐proportionally. Rapid and sustained inhibition of dipeptidyl peptidase‐4 reached 91–93% across linagliptin doses at steady state. At the end of the 24‐h dosing interval, inhibition was still high (82–90%). There were marked increases in plasma glucagon‐like peptide‐1 after 28 days of dosing. Compared to placebo, all linagliptin doses resulted in statistically significant decreases of the area under the glucose curve following a meal tolerance test on day 29, that is, 24 h after the last study drug intake. After 28 days of treatment with linagliptin the placebo‐corrected mean change in haemoglobin A1c (HbA1c) (median baseline 7.0%) was −0.31% (2.5‐mg dose), −0.37% (5‐mg dose) and −0.28% (10‐mg dose). The frequency of adverse events was similar for linagliptin (31%) and placebo (34%). There were no notable safety concerns. Conclusions: Linagliptin administration led to attenuation of postprandial glucose excursions and, despite a low HbA1c at baseline, statistically significant reductions in HbA1c after only 4 weeks of treatment. Linagliptin had a safety and tolerability profile similar to placebo in T2DM patients.