Background: Superior sulcus (Pancoast) tumors are preferably treated with neoadjuvant chemoradiotherapy and surgical resection. Induction chemoradiation increases the probability of a complete resection and local control, but is associated with an increased complication rate after surgery. Comorbidity, pulmonary reserve and response to induction treatment determine whether this demanding trimodality treatment can be advocated. Resections generally involve the upper lobe with contiguous thoracic wall including apical ribs (and T1 plexus root /partial vertebra if necessary), but even more extensive resections have been performed in fit patients. We evaluated the role of surgery as part of multimodality treatment in our institute. Methods: From registration databases all patients with Pancoast tumors, referred to our institute between 1995 and 2006, were identified. All patients without distant metastasis were selected for treatment with curative intent. The preferred treatment consisted of chemoradiation (66 Gy radiotherapy in fractions of 2,75 Gy with daily cisplatinum 6 mg/m 2 ) followed by resection. Presenting symptoms, performance status, pulmonary function, clinical and pathological tumor stage, (response to) treatment, and survival were reviewed retrospectively. Survival analysis was performed using the Kaplan-Meier method. Results: From 1995 to 2006, 85 patients with Pancoast tumors, 57 men and 28 women, were referred to our institute. The median follow-up was 15 months (2-123). Mean age was 57 years (32-82). After additional diagnostic staging, 25 patients had stage IIB (29%), 7 stage IIIA (8%), 32 stage IIIB (38%) and 21 stage IV (25%). Of those patients presenting with stage II or III disease, 22 underwent a resection after induction treatment. The 2-and 5-years overall survival was 70% and 37%, respectively. All these resections were complete and local recurrences were not observed. In 13 out of 22 patients a pathologic complete tumor response (pCR) was found after induction treatment. pCR was a significant prognostic factor for survival (5 year survival of 50% vs. 17%). In most cases, pathologic response was not evident from radiologic imaging (restaging with CT or MRI). The morbidity of surgery after induction treatment was acceptable: 6 pneumonias, 1 tracheoesophageal fistula, 1 chylothorax, 1 complete atelectasis, 1 postoperative bleeding. There were no fatal toxicities or treatment-related mortalities. Forty-two patients were rejected for resection due to comorbidity/inoperability (n=25), irresectability or insufficient response on induction treatment. Nineteen of these patients received concurrent chemoradiation (2-and 5-year survival 39% resp. 15%) and 23 patients were treated with radiotherapy, chemotherapy or both sequentially (2-and 5-year survival 18% resp. 5%). Local recurrence/progression occurred in 13 out of 42 patients, six of these patients had distant metastases as well. Conclusions: Surgical resection following induction chemoradiation is associated with excellent local control and ac...