Dipeptidyl peptidase-4 (DPP-4) inhibitors are oral antidiabetic agents that hold the potential of slowing the progress of type 2 diabetes mellitus. Their long-term safety is still a subject of debate. A systematic review of randomized, controlled trials was undertaken to comprehensively profile the safety of chronic treatment of type 2 diabetes mellitus with DPP-4 inhibitors. We searched data sources including MEDLINE, CENTRAL, publishers' and manufacturers' databases. Eligible trials were double-blind, randomized, placebo or active-controlled trials with ≥18 weeks duration in patients with type 2 diabetes reporting safety outcomes. Meta-analysis was performed separately for trials in which the control group received placebo (44 studies), another gliptin (3 studies) and any other antidiabetic drug (20 studies). Risk ratios with 95% confidence intervals were computed using a Mantel-Haenszel fixed-effect model for general safety outcomes, hypoglycaemia and adverse events by system organ class. Of 307 publications retrieved, 67 randomized, controlled trials met the eligibility criteria and were included in this review (4 alogliptin, 8 linagliptin, 8 saxagliptin, 20 sitagliptin, and 27 vildagliptin trials). Adverse events with gliptin treatment were at placebo level (relative risk (RR) 1.02 [0.99, 1.04]). No increased risk of infections was detectable (RR 0.98 [0.93, 1.05] compared to placebo and 1.02 [0.97, 1.07] compared to other antidiabetic drugs). Asthenia (RR 1.57 [1.09, 2.27]) as well as cardiac (RR 1.37 [1.00, 1.89]) and vascular disorders (RR 1.74 [1.05, 2.86] for linagliptin) emerged as adverse events associated with DPP-4 inhibitor treatment. The risk of hypoglycaemia was low with DPP-4 inhibitor treatment (RR 0.92 [0.74, 1.15] compared to placebo, RR 0.20 [0.17, 0.24] compared to sulphonylureas) in the absence of sulphonylurea or insulin co-therapy, but significantly elevated for combination therapy of sulphonylurea or insulin with sitagliptin or linagliptin (RR 1.86 [1.46, 2.37] compared to placebo). A large body of data supports the long-term safety of gliptin treatment and refutes an increased risk of infections. Further research is needed to clarify a possible link to asthenia, cardiac and vascular events. For combination therapy with insulin or insulin secretagogues, a careful choice of the agent used may limit the risk of hypoglycaemia.