ABSTRACT: BACKGROUND:Preoperative low left ventricular ejection fraction (LVEF) has been reported as an independent risk factor for in-hospital mortality. However, there were few studies evaluating the long-term mortality in these patients. We, therefore, conducted this study to investigate long-term outcomes of surgery on patients with LVEF?35% undergoing a broad range of cardiac procedures. METHODS:We performed a retrospective cohort study in 510 patients from January 1, 2007 to September 1, 2019. These patients were divided into survival group (n?=?386) and non-survival group (n?=?124). The multivariate Cox analysis was used to estimate the risk factors for survival. In Cox analysis, ?-blockers were indicated to be associated with long-term mortality. To further address bias, we derived a propensity score predicting the function of ?-blockers on survival, and matched 52 cases to 52 controls with similar risk profiles. RESULTS:Patients were followed for a median period of 24?months (interquartile range: 11-44?months). Multivariate Cox regression analysis indicated that the non-survival group had higher weight, higher EuroSCORE, more smoking patients, longer time of cardiopulmonary bypass (CPB), more intra-aortic balloon pump (IABP) use, and more patients who always used ?-blocker (HR: 2.056, 95%CI:1.236-3.420, P?=?0.005) compared with survival group. After propensity matching, the group which always used ?-blocker showed higher rate of all-cause death compare with the control group (61.54% vs 80.77%, P?=?0.030). CONCLUSIONS:The risk factors for long-term survival were weight, EuroSCORE, smoking, CPB, IABP, always used ?-blockers in patients with LVEF?35%. The discharge prescription of ?-blocker should be cautiously administrated in those patients.