colleagues on "Arterial complications of pancreatitis: Diagnostic and therapeutic role of radiology" in this issue of Pancreas (pp. 687-692). Savastano et al. describe the results of operative angiographic therapy of nine patients with arterial complications of pancreatitis. Six patients had pseudoaneurysms associated with pancreatitis, one had postoperative bleeding from the splenic artery, and two had pseudoaneurysms associated with the pancreaticoduodenal artery but none associated with pseudocyst. There was one death among the nine patients: the patient who experienced postoperative bleeding from the splenic artery.The authors rightly emphasize the important role angiography holds in the diagnosis of arterial complications of pancreatitis. However, they recommend limiting angiographic embolization therapy to patients having pseudoaneurysms of the pancreatic vasculature in the absence of pseudoc! st, except when the patient is actually bleeding, and then only for temporary control.This writer's own experience does not support this view. In Case 8, in which a pseudoaneurysm was associated with a pseudocyst, embolization provided definitive therapy for the pseudoaneurysm. While the associated pseudocyst required surgery, including a cystojejunostomy only two weeks after embolization of the pancreaticoduodenal artery, no bleeding occurred then or during a follow-up period of two years. Had the authors chosen to embolize Cases 1 , 2, 3 , 6 , and 7, they might have avoided two pancreaticoduodenectomies, at the least, and perhaps three distal pancreatectomies and splenectomies. In our experience, and that of others, embolization of the pseudoaneurysms has been definitive therapy for the pseudoaneurysm, whether the aneurysm is associated with a cyst or not. Once the pseudoaneurysm is angiographically controlled, the associated pseudocyst does not need to be resected but can be drained by cystogas-trostomy or cystojejunostomy . In some instances the pseudocyst may resolve spontaneously and not require operative therapy either. Resective pancreatic surgery is infrequently required for pancreatic control of bleeding from pseudoaneurysms when angiographic embolization is aggressively pursued, whether or not the pseudoaneurysm is associated with a pseudocyst.