OBJECTIVES The aim of this study was to compare 5-year rates of overall death, cardiac-related death, myocardial infarction, repeat revascularization, stroke and new occurrence of postoperative renal failure in a large cohort of patients with coronary disease, treated with on- or off-pump coronary artery bypass grafting (CABG). METHODS Two propensity score-matched cohorts, each of 560 patients, undergoing isolated surgical coronary revascularization at the regional public and private centres of Emilia-Romagna region (Italy) over the period 1 January 2003 – 31 December 2013, were used to compare long-term outcomes of on-pump CABG (6711 patients) and off-pump CABG (597 patients). RESULTS The matched on-pump group received significantly more bypass grafts than the matched off-pump group (2.4 ± 1.1 vs 1.6 ± 0.9, P < 0.0001). The on-pump group reported statistically significant lower cardiac-related mortality. There was a trend towards higher overall mortality and the need for repeat revascularization procedures in the off-pump group. No difference was found for myocardial infarction, stroke or new occurrence of postoperative renal failure between groups in the follow-up. The multivariate analysis of significant predictors of mortality in the overall population confirmed that the off-pump revascularization strategy was an independent predictor of death at long-term follow-up. On-pump CABG reported significantly better results in terms of mortality in the subgroups of patients with a depressed left ventricular ejection fraction and in patients with three-vessel disease. CONCLUSIONS In patients undergoing elective isolated CABG, on-pump strategy conferred a long-term survival advantage compared with off-pump strategy, particularly for patients with more extensive coronary disease. No benefits were found in terms of reduction of postoperative morbidity with the off-pump strategy. On-pump surgery should be the preferred revascularization technique, and off-pump surgery reserved for patients for whom the perioperative risk of cardiopulmonary bypass is greater than the risk of a less complete coronary revascularization.