Abstract OBJECTIVES Recent studies have once again brought into focus the long-term survival following off-pump coronary artery bypass grafting (OPCAB) compared with conventional on-pump coronary artery bypass grafting surgery (ONCAB). The aim of this study was to compare the long-term risk-adjusted survival rates in patients undergoing coronary artery bypass grafting (CABG) using these 2 techniques. METHODS We undertook a propensity score-matched analysis of 10 293 patients who underwent CABG at our single institution between 2000 and 2016. A logistic regression model was fitted using 14 covariates and their 2-way interactions to calculate an estimated propensity score [area under curve (AUC) 0.69], from which 1:1 nearest neighbour matching was performed. Patient survival was assessed using the Kaplan–Meier method and log-rank test. RESULTS Of the total cohort, 8319 patients had ONCAB and 1974 had OPCAB. Prior to matching, the OPCAB group had marginally higher EuroSCORE [3.7 ± 2.7 vs 3.5 ± 3, median (interquartile range) 3 (2–5) vs 3 (2–5), P = 0.016] and significantly lower average number of grafts per patient (2.39 ± 0.72 vs 2.75 ± 0.48, P < 0.001). Post-matching distributions between OPCAB and ONCAB showed a substantial improvement in balance in preoperative patient characteristics. The 2 surgery groups differed significantly in survival (P < 0.001). OPCAB demonstrated improved long-term survival at 10 years [84.8%, 95% confidence interval (CI) (82.7–86.9%) vs 75.8%, 95% CI (73.4–78.2%)] and 15 years [65.4%, 95% CI (61.4–69.6%) vs 58.5%, 95% CI (54.9–62.3%)]. Results of sensitivity analysis for 1:2 and 1:3 matched data were in concordance with these findings of survival. CONCLUSION At our institution, selected patients who underwent OPCAB had lower in-hospital morbidity and improved long-term survival when compared with a matched population of ONCAB patients.