Objectives: Experience of hypnosis in gastrointestinal (GI) endoscopy is scarce in children. Our aims were to assess the rate of successful GI endoscopy performed using hypnosis alone or in combination with midazolam, with or without additional equimolar mixture of oxygen and nitrous oxide (EMONO), and to identify predictive factors of successful endoscopy in children. Methods: This prospective single-centre study included children older than 6 years requiring a diagnostic esophagogastroduodenoscopy (EGD) or rectosigmoidoscopy. Ericksonian hypnosis was performed alone or in combination with midazolam, with or without additional EMONO. Successful endoscopy was defined by a complete and well-tolerated procedure. Levels of satisfaction of the endoscopist, nurse, and patient were assessed. Results: One hundred forty children [70 boys, median age: 12 years (Q1–Q3: 9–14)] were included over a 14-month period. They underwent EGD in 51.4% (n = 72) and rectosigmoidoscopy in 48.6% (n = 68) of cases. EMONO and midazolam were combined with hypnosis in 136 cases (97.1%). Successful endoscopy rate reached 82.9%. The procedure was interrupted due to poor tolerance and was rescheduled under general anaesthesia in 11 patients (7.9%). Predictive factors for successful endoscopy were older age (13 vs. 8 years, OR: 1.34, CI 95% [1.10–1.62], p = 0.003) and type of endoscopy (EGD vs. rectosigmoidoscopy, OR: 16.34 [2.14–124.68], p = 0.007). A good cooperation of the patient was reported by the endoscopist and the nurse in 88.4 and 86.9% of cases, respectively. Ninety-two per cent of patients mentioned that the procedure went well. Conclusions: Our study suggests that hypnosis combined with EMONO and/or midazolam is of additional value to perform diagnostic EGD or rectosigmoidoscopy in children older than 6 years without systematic need for general anaesthesia.