Ideal sedation for endoscopic submucosal dissection (ESD) aims to satisfy both the endoscopist and patient. However, previous studies show that a satisfactory procedure for the endoscopist does not equal higher patient satisfaction. This study attempted to find a sedation protocol that is able to increase patient satisfaction during propofol-based sedation by adding low-dose midazolam as premedication.Seventy-two adult patients were randomly allocated to receive either 0.02 mg/kg midazolam (Midazolam Group) or placebo (Control Group) as premedication before ESD. Sedation was done by targeting Modified Observer's Assessment of Alertness/Sedation (MOAA/S) scale of 3 or 4 with continuous propofol infusion and bolus doses of fentanyl. Satisfaction scores of the endoscopists and patients, and whether the patient was willing to receive the same sedation method in the future was assessed. Interim analysis was done after enrollment of 50 % of patients.This study was prematurely terminated when interim analysis showed that patients willing to receive the same sedation method in the future were significantly lower in the Control Group compared to the Midazolam Group (P = 0.001). There was no difference in sedation time, procedure and recovery time, drug requirements and adverse events between the two groups. Endoscopist and overall patient satisfaction scores, patient pain scores and degree of recall were also similar between groups.A small dose of midazolam given as premedication before propofol-based sedation is able to reduce patient reluctance to repeat the same procedure in the future, without affecting procedural performance, recovery time or endoscopist satisfaction.