Background: Achieving post-anesthesia discharge criteria after surgery or outpatient procedures does not mean that the patient has regained all his or her faculties, such as driving. Although mandated by many clinical guidelines, there is no evidence that escort-drivers reduce the risk of traffic accidents after deep sedation. The purpose of this study was to evaluate that hypothesis that driving performance as measured using a driving simulation would not differ between patients who had undergone deep sedation for gastrointestinal endoscopy meeting discharge criteria and their escorts. Methods: This prospective study included patients scheduled for ambulatory gastrointestinal endoscopy under deep propofol sedation (patient group) and their escorts (escort group). Driving performance of escorts and patients (when discharge criteria were met) was assessed using a driving simulator. Results: 30 patients and their escorts were included. Patients crossed the midline significantly more frequently than escorts (3 [2-4] (median [IQR]) and 2 [1-3] crossings, respectively, p=0.015]. Patients were speeding for a higher proportion of the distance traveled compared with escorts (37 (20)% (mean (SD)) and 24 (17)% in patients and escorts, respectively, p = 0.029). There were no significant differences between groups in other simulation parameters. Conclusions: The ability to stay within the traffic lanes, as measured by the number of midline crossing during a simulated driving performance, is impaired in patients who meet discharge criteria after gastrointestinal endoscopy under deep sedation compared with their escorts. This finding does not support a practice of allowing patients to drive themselves home after these procedures.