BACKGROUND The putative relationship between pancreas divisum and clinical pancreatitis continues to be controversial. Indications for surgical or endoscopic manipulations of the accessory duct ampulla are unclear, and the relative merits of surgical sphincteroplasty, endoscopic sphincterotomy, and accessory duct stenting have not been established. METHODS Thirty-seven patients with documented pancreas divisum and acute pancreatitis identified by readily available clinical criteria were entered into a prospective five-year study of the value of surgical sphincteroplasty in preventing recurrent episodes of acute pancreatitis and eliminating "pancreatic pain." A broader therapeutic perspective was developed by means of an extensive analysis of existing surgical and endoscopic literature. RESULTS Surgical accessory duct sphincteroplasty was successful in 83.8 percent of our patients in the long-term prevention of recurrent acute pancreatitis, but was significantly less successful in improving "pancreatic pain" (p < 0.001). Collected surgical results are superior to reported outcomes from endoscopic accessory papillotomy with regard to relief of "pancreatic pain," rate of restenosis, and generation of procedure-specific complications. Prolonged stenting of the accessory duct cannot be recommended. CONCLUSIONS Patients with pancreas divisum and recurrent acute pancreatitis who will benefit from therapy can be identified by clinical means. Surgical sphincteroplasty provides superior long-term results compared to endoscopic sphincterotomy.