Background A recent American Society for Gastrointestinal Endoscopy Technology Status Evaluation Report recommended comparative studies of water-aided colonoscopy methods to refine the optimal insertion technique. Objective Air insufflation (AI), water immersion (WI), and water exchange (WE) were compared head-to-head to test the hypothesis that WE produces the least insertion pain. Design Patient-blinded, prospective, randomized, controlled trials. Setting Two community hospitals in Italy. Patients First-time diagnostic or screening colonoscopy in unsedated patients with the option of on-demand sedation. Intervention Colonoscopy with AI, WI, or WE. Main Outcome Measurements Real-time maximum insertion pain (0 = none, 10 = worst). To avoid interventional bias, the timing of recording was at the discretion of the nurse assistant. Adjunct measures were implemented to ensure patient perception of minimal discomfort. Recalled pain and patients’ guess of insertion methods were recorded after colonoscopy. Results Results were merged for 576 randomized patients. Correct patient guesses lower than 33% confirmed adequate blinding. Significant correlation (Pearson coefficient 0.6, P < .0005) between real-time and recalled pain provided internal validation of the former as the primary outcome. Real-time pain (95% confidence interval [CI]: AI, 4.1 [3.7-4.5]; WI, 3.5 [3.0-3.9]; and WE, 2.5 [2.2-2.9] [P < .0005] was the lowest in the WE group. The proportions of patients completing unsedated colonoscopy based on the assigned methods were significantly different (WE, 74.7% vs WI, 62.4%; P = .009; vs AI, 65.3%; P = .04). WE required the least implementation of adjunct maneuvers. Limitations Unblinded colonoscopists. Conclusion The current findings with an internally validated primary outcome in adequately blinded patients support the hypothesis that WE is superior to WI in attenuating real-time insertion pain and enhancing completion of unsedated colonoscopy.