Background: Current guidelines recommend monitoring the anesthetic depth of sedation during respiratory endoscopy by using clinical scales despite their subjective nature and the potential change in the level of sedation caused by frequent stimulation. Monitoring by means of the bispectral index (BIS) has shown its utility in reducing the use of drugs and their adverse events in general anesthesia, but evidence in prolonged sedation is insufficient. Objective: Our objective was to evaluate BIS in patients undergoing endobronchial ultrasound (EBUS). Methods: A randomized cohort study of 90 patients with mediastinal lymph node involvement and/or lung or mediastinal lesions for whom EBUS was indicated, comparing the modified observer's assessment of alertness/sedation scale clinical evaluation (n = 45) versus the BIS evaluation (n = 45) of sedation with propofol-remifentanil, was conducted in order to evaluate the clinical parameters, doses used, adverse events, and tolerance of the procedure. Results: We found a shorter waking time and a significantly lower dose of total propofol in the BIS group. Significantly fewer overall adverse events were recorded in the BIS group and included desaturation, hypotension, and bradypnea. Tolerance was better in the BIS group. No significant differences were found in terms of cough, memory of the procedure, or the level of difficulty of EBUS on the part of the pulmonologists. Conclusions: BIS monitoring of sedation in EBUS makes it possible to reduce the dosage of propofol, thereby shortening the waking time and reducing adverse events. This form of monitoring should be taken into consideration in the future for systematic use in prolonged sedation, as in the case of EBUS.