Abstract OBJECTIVES The aim of this study was to analyse the outcomes of an international cohort of patients affected by Masaoka stage III thymic epithelial tumours with vascular involvement and treated by surgery. METHODS Study design was the observational multicentre retrospective cohort study. Data were extracted from the European Society of Thoracic Surgeons thymic database; additional variables were collected. Inclusion criteria were as follows: stage III (Masaoka–Koga) thymic epithelial tumours; surgery with radical intention; clinical or pathological great vessels involvement; and radiologically suspected or diagnosed intraoperatively. Outcome items were analysed. RESULTS Sixty-five patients submitted to surgery from 2001 to 2017 fulfilled inclusion criteria. Thymoma and thymic carcinoma patients did not differ for demographics and clinical characteristics. The majority of great vessel treated were superior vena cava or innominate veins (72.3%). Eleven patients (16.9%) had postoperative cardiopulmonary complications; vascular stenosis was observed in 3 patients (4.6%). The multivariable Cox analysis for disease-free survival showed an increased hazard of recurrence for thymic carcinoma (hazard ratio = 3.59; 95% confidence interval: 1.66–7.78, P = 0.001). The 1-, 3-, 5- and 10-year overall survival rates were 0.86, 0.84, 0.81, and 0.53, respectively. There was no significant difference in overall survival according to resection status or between thymoma and thimic carcinoma. The univariable Cox regression model did not show an increased hazard of death for myasthenic patients considering all resection status and for patients who received neoadjuvant therapy. CONCLUSIONS We observed that clinical outcomes of patients treated for stage III thymic epithelial tumours with vascular involvement are satisfactory suggesting to increase the confidence in dealing with these complex surgeries. Complete resection should be achieved, even though extensive vascular reconstructions are required.