Background: Celiac plexus neurolysis, a chemical splanchnicectomy of the celiac plexus, is used to treat pain caused by pancreatic cancer. Most commonly, celiac plexus neurolysis is performed percutaneously under CT or fluoroscopic guidance, but can also be performed with EUS. The aim of this study was to prospectively assess the efficacy of EUS celiac plexus neurolysis in the management of pain caused by pancreatic cancer. Methods: In this prospective study conducted in a community-based referral hospital, 58 patients with painful and inoperable pancreatic cancer were evaluated at 8 observation points before and after EUS celiac plexus neurolysis for up to 6 months. The following data were collected: age, gender, tumor location, vascular invasion, adjuvant therapy, and laboratory tests including prothrombin time, and complete blood counts were obtained at baseline (before EUS celiac plexus neurolysis); pain scores, morphine use, and adjuvant therapy were assessed at each observation. Results: Pain scores were lower (p = 0.0001) 2 weeks after EUS celiac plexus neurolysis, an effect that was sustained for 24 weeks when adjusted for morphine use and adjuvant therapy. Forty-five of the 58 patients (78%) experienced a decline in pain scores after EUS celiac plexus neurolysis. Chemotherapy with and without radiation also decreased pain after EUS celiac plexus neurolysis (p = 0.002). Procedure-related transient abdominal pain was noted in 5 patients; there were no major complications. Conclusions: EUS celiac plexus neurolysis is safe and controls pain caused by unresectable pancreatic cancer. (Gastrointest Endosc 2001;54:316-24.)