Abstract Aims To evaluate the prognostic value of global and regional quantitative [15O]H2O positron emission tomography (PET) perfusion. Methods and results In this retrospective study, 648 patients with suspected or known coronary artery disease (CAD) who underwent [15O]H2O PET were followed for the occurrence of death and myocardial infarction (MI). Global and regional hyperaemic myocardial blood flow (hMBF) and coronary flow reserve (CFR) were obtained from [15O]H2O PET. During median follow-up of 6.9 (5.0–7.9) years, 64 (9.9%) patients experienced the composite of death (36–5.6%) and MI (28–4.3%). Impaired global hMBF (<2.65 mL/min/g) and CFR (<2.88) were both significant prognostic factors for death/MI after adjusting for clinical characteristics (both P < 0.001). However, after adjusting for clinical parameters and the combined use of hMBF and CFR, only hMBF remained an independent prognostic factor (P = 0.04). For regional perfusion, both impaired hMBF (<2.10 mL/min/g) and CFR (<2.07) demonstrated prognostic value for events (both P < 0.001). Similarly, after adjusting for clinical characteristics and combined use of hMBF and CFR, only hMBF had independent prognostic value (P = 0.04). The combination of global and regional perfusion did not improve prognostic performance over either global (P = 0.55) or regional perfusion (P = 0.37) alone. Conclusion Global and regional hMBF and CFR were all prognostic factors for death and MI. However, for both global and regional perfusion, hMBF remained the only independent prognostic factor after adjusting for the combined use of hMBF and CFR. Additionally, integrating global and regional perfusion did not increase prognostic performance compared to either regional or global perfusion alone.