Background: Approaches of thoracoscopic thymectomy for myasthenia gravis (MG) are debatable. The subxiphoid approach is widely utilized recent years for its better visualization of the anterior mediastinum. In the present study, we compared perioperative outcomes and mid-term effects of the extended thymectomy for MG between the subxiphoid approach and the routine right-thoracic approach. Methods: One hundred and thirty-one MG patients treated with thoracoscopic extended thymectomy were analyzed. Among them, 68 patients were operated on via the subxiphoid approach and the other 63 via the conventional right-side unilateral approach. The patient outcomes were retrospectively reviewed and evaluated. Mid-term clinical outcome was assessed according to the classification system proposed by the Myasthenia Gravis Foundation of America (MGFA). Clinical efficacy and variables influencing outcome were evaluated by the Kaplan-Meier method and Cox proportional hazards regression analysis. Results: Compared with the right thoracic approach, the duration of the procedure via the subxiphoid approach was significantly shorter (P=0.035), the rates of total thymectomy were higher (P=0.028), and the pain scores on postoperative days 1, 3, and 7 were significantly lower (P<0.001, P<0.001, and P=0.03, respectively). A total of 112 patients with MG were followed up. The subxiphoid approach group reported higher rates of complete stable remission (CSR) and effective treatment of MG, although these differences were not statistically significant (Z=−0.484, P=0.627). By multivariate Cox proportional hazards modes analysis, the chance of CSR was significantly increased when age <40 (OR: 2.