Aims: Left atrial appendage (LAA) occlusion for Atrial fibrillation (AF) is an invasive therapy to reduce stroke incidence and death. Cardiac rehabilitation based on exercise training has become part of the standard care for patients with heart disease. Studies are needed to evaluate the effects of cardiac rehabilitation on exercise capacity, cardiac function, quality of life (QoL), residual permeability, device endothelialization and safety in AF patients with LAA occlusion. Methods: A single-center, three-month randomized controlled trial comparing an exercise-based cardiac rehabilitation program with a control group was conducted. AF patients underwent LAA occlusions were randomized to exercise-based cardiac rehabilitation (rehabilitation group) versus usual care (control group). All patients received assessments of echocardiography, cardiac CT, scale survey, physical function and exercise capacity. All analyses were performed using SPSS 19.0. Results: Finally, 33 patients in control group and 30 patients in rehabilitation group entered final analysis. The six-minute walk distance, handgrip strength, leg strength and left ventricular systolic function increased significantly after the study period in the rehabilitation group compared with baseline, which have significant interactions with control group. Significant differences between groups were found for the mental and physical component summary scales on SF-36, also in all of the 8 subscales. The incidence of peri-device leak and incomplete endothelialization in cardiac rehabilitation group were both significantly lower than that in control group. There was no significant difference in the incidence of adverse events between the two groups. Conclusion: Cardiac rehabilitation could promote device endothelialization and reduce peri-device leak in AF patients with LAA occlusion, which also has benefits in QoL, exercise capacity, physical function and systolic left ventricular function.