Abstract Aims The accuracy and prognostic value of stress perfusion cardiac magnetic resonance (CMR) are established in coronary artery disease (CAD) patients. Because myocardial contrast kinetics may be altered after coronary artery bypass graft (CABG), most studies excluded CABG patients. This study aimed to assess the prognostic value of vasodilator stress perfusion CMR in CABG patients. Methods and results Consecutive CABG patients referred for stress CMR were retrospectively included and followed for the occurrence of major adverse cardiovascular events (MACE) including cardiovascular (CV) death or non-fatal myocardial infarction (MI). Cox regression analyses were performed to determine the prognostic association of inducible ischaemia and late gadolinium enhancement (LGE) by CMR. Of 866 consecutive CABG patients, 852 underwent the stress CMR protocol and 771 (89%) completed the follow-up [median (interquartile range) 4.2 (3.3–6.2) years]. There were 85 MACE (63 CV deaths and 22 non-fatal MI). Using Kaplan–Meier analysis, the presence of inducible ischaemia identified the occurrence of MACE [hazard ratio (HR) 3.52, 95% confidence interval (CI): 2.27–5.48; P < 0.001] and CV death (HR 2.55, 95% CI: 1.52–4.25; P < 0.001). In a multivariable stepwise Cox regression including clinical characteristics and CMR indexes, the presence of inducible ischaemia was an independent predictor of a higher incidence of MACE (HR 3.22, 95% CI: 2.06–5.02; P < 0.001) and CV death (HR 2.15, 95% CI: 1.28–3.62; P = 0.003), and the same was observed for LGE (both P = 0.02). Conclusion Stress CMR has a good discriminative prognostic value in patients after CABG, with a higher incidence of MACE and CV death in patients with inducible ischaemia and/or LGE.