In an eight-month period, four patients in our peritoneal dialysis program developed acute pancreatitis, an incidence significantly higher than that in our hemodialysis program. Diagnosis was difficult since the symptoms of pancreatitis were similar to those of peritoneal dialysis-associated peritonitis. Further difficulties in diagnosis were due to unreliability of serum amylase levels and "routine" ultrasound examinations in suggesting the presence of pancreatitis. Computerized tomography performed in three patients showed enlarged, edematous pancreata with large extrapancreatic fluid collections in all cases. Two patients died, one directly due to complications of pancreatitis. One patient was changed to hemodialysis and showed clinical and radiologic resolution of his pancreatitis. One patient remains on peritoneal dialysis but has now had four attacks of acute pancreatitis. No patient had classic risk factors for development of pancreatitis. Review of patient histories showed no common historical factors except for renal failure itself, peritoneal dialysis, peritonitis, catheter surgery, and hypoproteinemia. It is possible that metabolic abnormalities related to absorption of glucose and buffer from dialysate or absorption of a toxic substance present in dialysate, bags, or tubing can cause pancreatitis in patients on peritoneal dialysis. We feel that a diagnosis of pancreatitis should be considered when peritoneal dialysis patients present with abdominal pain, particularly if peritoneal fluid cultures are negative or if patients with positive cultures do not have prompt resolution of symptoms with appropriate antibiotic therapy.