Objective. Different regimens are used for sedation during endoscopic retrograde cholangiopancreatography (ERCP). Our objectives were to compare safety, ease of treatment, recovery, and patients’ experiences using patient-controlled sedation (PCS) with propofol, nurse anesthetist-controlled sedation (ACS), or the department’s standard of care, midazolam given by the procedure team (control group). Material and methods. The study included 281 adults in 301 procedures. The PCS group (n = 101) delivered bolus doses of 5 mg of propofol according to their need for sedation. The ACS group (n = 100) had 2–8 mg/kg/h of propofol infused, with the target for sedation being level 3 of the Observer’s Assessment of Alertness/Sedation (OAA/S) scale. The control group was given 2–3 mg of midazolam for induction and additional 1 mg if required. Results. PCS and ACS increased the ease of the procedure and reduced the number of sedation failures compared to midazolam sedation (ACS n = 0; PCS n = 4; midazolam n = 20). The ACS group had more deeply sedated patients (OAA/S level 2), desaturation, and obstructed airways than the PCS and midazolam groups. Time to full recovery (Aldrete score ≥9) was shortest following PCS. PCS resulted in the least fatigue and pain after the procedure. Patients’ preference for PCS and ACS was the same. Conclusion. PCS with propofol is superior to midazolam and comparable to ACS. PCS resulted in a rapid recovery, fewer respiratory events, and was almost as effective as ACS in ensuring a successful examination.