Background: Dexmedetomidine could be an ideal adjuvant to propofol during gastrointestinal endoscopy because it provides both analgesia and sedation without respiratory depression. This study investigates the effect of different doses of dexmedetomidine on the median effective concentration of propofol during gastrointestinal endoscopy. Methods: 90 adult patients were randomly assigned to Group Control , Group DEX0.5 (0.5 μg/kg dexmedetomidine), or Group DEX1.0 (1.0 μg/kg dexmedetomidine) . Anaesthesia during endoscopy was implemented by plasma target-controlled infusion (TCI) of propofol with different doses of dexmedetomidine. TCI concentration of the first patient for each group was 2.5 μg/ml and the consecutive adjacent concentration gradient was 0.5 μg/mL. EC50 of TCI propofol for gastrointestinal endoscopy was determined by using the modified Dixon’s up-and-down method. Cardiovascular variables were also measured. Results: EC50 of TCI propofol and 95% confidence interval (CI) for gastrointestinal endoscopy were, 3.77 (3.48-4.09), 2.51 (2.27-2.78) and 2.10 (1.90-2.33) μg/mL in Group Control, Group DEX0.5 and Group DEX1.0. The average percent change from baseline in HR was 2.8 (8.9), -7.4 (7.7) and -10.5 (8.8) (P＜0.001), and the average percent change from baseline in MAP was -10.6 [-24.7; 3.5], -9.5 [-29.2; 11.4] and -4.0 [-27.3; 15.5] (P = 0.034) in Group Control, Group DEX0.5 and Group DEX1.0, respectively. Conclusions: Dexmedetomidine reduced the EC50 of TCI propofol. A 0.5-1 μg/kg dexmedetomidine caused a decrease in HR without bradycardia. The decrease in dosage of propofol with increasing doses of dexmedetomidine caused more stable MAP. Dexmedetomidine is an ideal adjuvant drug to propofol during gastrointestinal endoscopy.