Introduction: Esophageal cancer (EC) causes high morbidity and mortality. Flexible bronchoscopy (FB) is useful for diagnosis, staging, and planning endoscopic intervention. We aimed to analyze the bronchoscopic characteristics of EC patients at the time of diagnosis, and its correlation with staging and mortality. Methods: Retrospective review of patients with a recent diagnosis of EC submitted to FB in our centre between 2006 and 2019. Results: we included 112 patients, with a median age of 64 years (41-96), 90.2% were male (n=102). 90% (n=101) had squamous cell carcinoma, 6.3% adenocarcinoma (n=7) and 3.6% undifferentiated carcinoma (n=4). Of squamous cell, 11.6% were stage II, 11.6% III, 30.4% IVA, 33.9% IVB and 6.3% unknown. Those with adenocarcinoma had stage III (n=1), IVA (n=2), IVB (n=3) or unknown (n=1). In 92.9% FB was performed after diagnosis (median of 27.5±59.8 days). 53.6% (n=60) had normal FB, 25.9% (n=29) had extrinsic compression and 20.5% (n=23) had tracheal/bronchial invasion, with fistula in 3. Invasion occurred mainly in proximal (17%, n=19), or middle trachea (7.1%, n=8). Invasion was associated with Stage IV (p<0.001), and with increased mortality, regardless of histology and stage (p<0.001). Treatment included chemo-radiotherapy in 26.8%, chemotherapy in 20.5%, neoadjuvant chemo-radiotherapy plus surgery in 14.3%, radiotherapy in 9.8%, surgery plus adjuvant chemo-radiotherapy in 0.9% and supportive care in 25%. The overall survival was 60.5% at 6 months and 40.1% at 12 months. Conclusion: FB is important for initial evaluation of EC patients, and is useful for prognosis, especially when there is invasion. It also identifies patients who benefit from endoscopic treatment, improving quality of life.