Background: Accepted techniques for pancreatic sphincterotomy include use of a needle knife or a pull sphincterotome to ablate the sphincter. There are no prospective studies comparing outcomes between both techniques. Aim: Compare post-ERCP pancreatitis rates among high-risk patients undergoing pancreatic sphincterotomy with the pull-sphincterotome versus needle-knife technique. Study Design: Prospective, randomized trial. Methods: Patients diagnosed with pancreatic sphincter hypertension at sphincter of Oddi manometry were randomized to undergo pancreatic sphincterotomy with a pull sphincterotome (followed by pancreatic stenting) or a needle knife over a pancreatic stent. Main Outcome Measurements: To compare post-ERCP pancreatitis rates between the pull-sphincterotome and needle-knife groups. Results: Forty-eight patients were randomized. Patient demographics and the incidence of patient and procedure risk factors for pancreatitis were similar in both treatment groups. The trial was stopped early after an interim analysis showed that post-ERCP pancreatitis was significantly higher among patients undergoing sphincterotomy with a pull sphincterotome than a needle knife (7/24 = 29% [95% CI 13-51] vs 0/24 = 0% [95% CI 0-12]; P = .01). No other complications were encountered. At a mean follow-up of 12 months (range 6-18 months), 60% of patients had complete symptom relief, 15% partial symptom relief, and 25% recurrent symptoms. Reintervention rates and clinical outcomes were not significantly different between the 2 groups. Limitations: Short duration of follow-up. Conclusions: Pancreatic sphincterotomy is safer in high-risk patients when performed with a needle knife over a pancreatic stent. © 2006 American Society for Gastrointestinal Endoscopy.