Introduction: The modified Gloucester comfort scale (GCS), although not formally validated, is routinely used as a quality assurance measure to assess patient discomfort during colonoscopy. We want to determine whether endoscopist or nurse derived GCS accurately reflects patient experience. Method: Prospectively collected data from 123 patients undergoing colonoscopies at a single centre from August to September 2019. Patients rated their comfort score using a 1=best to 5=worst ordinal scale that was compared with endoscopist and nurse assessed GCS from 1-5. Data collection was blinded for each assessor. Fleiss’ s kappa statistics was used to investigate inter-rater agreement of comfort scores. Patients were also asked to rate their overall satisfaction of the procedure. Results: Endoscopists had better comfort score agreement with nurses (κ =0. 36; minimal agreement) than with patients (κ =0. 26; no agreement). Nurse derived comfort sores had no agreement with patients (κ =0. 11). Overall inter-rater agreement was minimal with greater variability at worse levels of discomfort. There was no significant difference in Fentanyl and Midazolam dose between those patients who reported mild discomfort versus those with high discomfort. Patients undergoing their first colonoscopy anticipated higher discomfort levels compared to those who had prior colonoscopy examination (OR 2. 67 [95% CI 1. 38-5. 23]; p=0. 0036) but this did not influence final pain scores. The only factor that was significantly associated with patient pain score in multivariate analysis was gender, with male more likely to report higher scores (worse comfort) than female patients (OR 2. 67 [95% CI 1. 18-6. 05]; p=0. 018). Patient comfort scores weakly correlated with their satisfaction of the procedure (spearman coefficient=0. 29). Conclusion: Endoscopist and nurse derived GCS are not reliable measures of patient comfort. There is a need for a validated, easy-to-use method to assess patients’ pain during colonoscopy. Assessment tools that utilize objective bedside cues such as the nonverbal pain scale may be a better quality indicator.