Background The role of selenium (Se) in the management of type 2 diabetes mellitus (T2DM) remains unclear. We systematically assessed the effectiveness and safety of Se supplementation in adults with T2DM. Methods MEDLINE, EMBASE and the Cochrane Library were searched up to April 2018 for randomised controlled trials (RCTs) evaluating the effectiveness of Se against a comparator on DM‐related outcomes. Results Four RCTs (241 participants) were included. In individual RCTs, Se supplementation significantly reduced fasting insulin levels mean difference (MD) = −3.6 μIU mL−1; 95% confidence interval (CI) = −6.36 to −0.84; MD = −5.8 μIU mL−1; 95% CI = −9.23 to −2.37, homeostasis model of assessment‐estimated insulin resistance (HOMA‐IR) (MD = −1; 95% CI = −1.79 to −0.21; MD = −1.6; 95% CI, −2.58 to −0.62) and homeostasis model of assessment‐estimated B cell function (HOMA‐B) (MD = −13.6; 95% CI = −23.4 to −3.8; MD = −22.6; 95% CI = −36.39 to −8.81). No effects of Se were noted on most of the other outcomes of interest. None of the RCTs assessed the mortality, diabetes‐related complications, non‐high‐density lipoprotein (non‐HDL), blood pressure and health‐related quality of life. The impact on HDL and fasting plasma glucose (FPG) was ambiguous. Only one adverse event (nausea) was reported as a reason for discontinuing the intervention; however, among the studies, the reporting was not accurate. Furthermore, only one RCT reported increase in FPG level in the Se group (MD = 36.38 mg dL−1; 95% CI = 15.39–57.37). Conclusions Currently, there is no evidence to support the effectiveness of Se supplementation in the T2DM population.