The ASTER trial is cited often as evidence of the superiority of endosonography over mediastinoscopy, although it has several noteworthy limitations that nullify its conclusions. First, a median of three mediastinal nodal stations were assessed in the endosonography group, whereas a median of four mediastinal nodal stations were assessed in the mediastinoscopy arm, firmly establishing that mediastinoscopy lends itself to a more thorough evaluation of the mediastinum. Second, the ASTER trial has limited generalizability to clinical practice because it excluded patients with a radiographically normal mediastinum. Up to 40% of patients with NSCLC have a radiographically normal mediastinum,10 and practice guidelines recommend invasive mediastinal staging for all patients, except perhaps those with a peripheral clinical stage IA tumor.1,11 Third, the ASTER trial used a surrogate end point (sensitivity for N2/N3) as its primary outcome rather than a patient-centered outcome, such as survival. Surrogate end points can have the unintended consequence of potentially misinforming investigators, clinicians, patients, and other stakeholders about the safety and effectiveness of an intervention (or diagnostic test), so much so that surrogate end points were the subject of scrutiny of the US Food and Drug Administration’s accelerated drug approval process.12 Other important outcomes were either not measured (ie, sufficient tissue for molecular analyses, efficiency) or not different between endosonography and mediastinoscopy (ie, negative predictive value, survival, health-related quality of life, complications). Additionally, the ASTER trial used an antiquated clinical end point—futile thoracotomy defined by the finding of N2 disease, a T4 tumor, small cell lung cancer, or benign disease at the time of resection—that has no relevance to modern day thoracic oncologic care. Stage IIIA N2 can now be appropriately treated with adjuvant therapy,13 and practice guidelines do not consider intraoperative (or even preoperative) detection of N2 to be a contraindication to resection.1,11,14 Practice guidelines likewise recognize that technically resectable T4 tumors are appropriately treated with surgery,1 and no modern thoracic oncology center would consider such resections as futile. Pulmonary resection is also a legitimate treatment option for limited-stage small cell lung cancer,15 and lung nodules in patients at high risk for lung cancer often are appropriately managed through resection, even though some of these nodules will be benign.11,16 Finally, the ASTER trial used a clinically invalid experimental group by making the treatment arm essentially a series of progressive steps in invasive staging (EBUS/EUS followed by mediastinoscopy if endoscopic staging was negative) compared with mediastinoscopy alone. The authors provided multiple opportunities for successful staging in the intervention arm compared with only one opportunity, mediastinoscopy, in the control arm, fundamentally making an invalid comparison. The ASTER trial does not provide evidence for the superiority of endosonography over mediastinoscopy but does suggest that a strategy of endosonography results in a less thorough mediastinal evaluation than surgical staging.