In experimental stroke models pretreatment with the newly introduced antidiabetic agents, glucagon‐like peptide 1 receptor (GLP‐1R) agonists, has been shown to exert neuroprotective effects. Published evidence with regard to the effect of treatment with GLP‐1R agonists on the risk of stroke was evaluated. Data from prospective randomized placebo‐controlled trials up to October 2018 involving GLP‐1R agonists which reported cardiovascular outcomes as primary end‐points of efficacy and/or safety were meta‐analysed. Five eligible multicentre randomized placebo‐controlled trials (ELIXA, LEADER, SUSTAIN, EXSCEL and HARMONY) were included. The pooled analysis (n = 42 358) showed a significant reduction by 13% in the risk of total stroke from treatment with GLP‐1R agonists versus placebo (risk ratio 0.87, 95% confidence interval 0.78–0.98, P = 0.021) with no significant heterogeneity between trials (Q = 4.094, P = 0.393, I2 = 2.307%). When only fatal stroke was included (this applied for the ELIXA, LEADER, EXSCEL and HARMONY trials), active treatment was associated with a non‐significant reduction by 16% compared with placebo (risk ratio 0.84, 95% confidence interval 0.60–1.17, P = 0.29). The findings of this meta‐analysis support the evidence from earlier experimental studies calling attention to potential ‘stroke protective’ effects from treatment with GLP‐1R.