Abstract Background We sought to compare long-term follow-up of coronary artery bypass grafting (CABG) with percutaneous coronary intervention (PCI) in elderly patients with left main or multivessel disease, hypothesising that completeness of revascularisation and severity of coronary artery disease are predictors of adverse outcomes. Methods Patients aged ≥75 years with multivessel disease or left main disease who underwent PCI or CABG between 2012–2016 were included in this retrospective cohort study. Baseline characteristics from the index procedure were collected. Severity of coronary artery disease and completeness of revascularisation were assessed. Primary outcome was all-cause mortality, in addition we captured major adverse cardiac and cerebral events, bleedings, recurrent angina and new onset atrial fibrillation. Results A total of 597 patients were included. Median follow-up was 4 years (interquartile range 2.8–5.3 years). At baseline, patients in the PCI group more often had a previous medical history of CABG and more frequently underwent an urgent procedure compared with patients in the CABG group. Mortality at 5‑year follow-up was significantly higher in patients who underwent PCI compared with CABG (39.9% vs 25.4%, p < 0.001). Furthermore, acute coronary syndrome (ACS), repeat revascularisation and recurrent angina occurred more frequently after PCI, while occurrence of bleedings and new onset atrial fibrillation were more frequent after CABG. Neither completeness of revascularisation nor severity of coronary artery disease was a predictor for any of the outcomes. Conclusion Long-term mortality was higher in elderly patients with multivessel disease undergoing PCI compared with CABG. In addition, patients undergoing PCI had a higher risk of ACS, repeat revascularisation and recurrent angina.