Objective To improve the prognosis of severe acute pancreatitis, preventing infectious complications, particularly infected pancreatic necrosis, is important. The present study evaluated the efficacy of peritoneal lavage for improving the prognosis of patients with severe acute pancreatitis. Patients We retrospectively reviewed the cases of 23 consecutive patients with severe acute pancreatitis who were treated with peritoneal lavage. Results Peritoneal lavage was started within 72 hours after the initial onset of symptoms in 20 patients (87%). The duration of peritoneal lavage, which was significantly correlated with the number of prognostic factors according to the revised Japanese criteria, Ranson score and serum C-reactive protein level at the start of peritoneal lavage, was a median of seven (3-22) days. There were no adverse events associated with the peritoneal lavage. Eight patients (35%) concurrently underwent continuous regional arterial infusion. Five days after starting peritoneal lavage, the patients' clinical conditions significantly improved. Overall, the survival rate was 96%. One patient (4%) died due to rupture of a pseudoaneurysm of the splenic artery. Complications occurred in seven patients (30%). Infectious complications were observed in three patients (13%) (one patient developed infected pancreatic necrosis and bacteremia, and two patients developed bacteremia). Pseudocysts and pancreatic fistulas developed in five and one patient, respectively. The incidence of complications was lower in the patients receiving peritoneal lavage within 72 hours from the initial onset of symptoms than in the remaining patients (20% vs. 100%; p=0.005). Conclusion We speculate that peritoneal lavage reduces the mortality and incidence of complications in patients with severe acute pancreatitis.