<b><i>Background:</i></b> Necrotizing pancreatitis is a life-threatening presentation of acute pancreatitis. The mortality of 20-80% initially depends on the persistence of organ failure and systemic inflammatory response syndrome (SIRS) and, in the later course of the disease, on secondary infection of the necrosis. The questions whether prophylactic antibiotics aiming to prevent this infection should be administered and which antibiotic is the best to use, as well as the problem of fungal infection under antibiotic treatment are still intriguing and insufficiently solved. <b><i>Methods:</i></b> A search of the literature using PubMed was carried out, supplemented by a review of the programmes of the Digestive Disease Week (DDW) and the United European Gastroenterology Week (UEGW). <b><i>Results:</i></b> Despite the widely practised prophylactic antibiotic administration in severe pancreatitis, no evidence for the benefit of this strategy exists. One of the drawbacks might be a tendency for disastrous fungal infection under prophylactic antibiotics. Bacterial translocation from the gut in the second week after the onset of symptoms is the major source for infection of pancreatic necrosis and provides a clear indication for antibiotic treatment. However, routine fine-needle aspiration for a calculated antibiotic therapy cannot be recommended, and all other tests offer only indirect signs. Important factors such as enteral versus parenteral feeding and the method of necrosectomy are mostly neglected in the trials but seem to be essential for the outcome of the patient. <b><i>Conclusions:</i></b> Even though most meta-analyses including the newer double-blind, placebo-controlled trials on prophylactic antibiotics showed no beneficial effects in the prevention of infection of necrosis and/or outcome of the patients, this strategy is still widely used in clinical routine. Since nearly all trials published so far show systematic problems (i.e. inaccurate definition of the severity of the disease, poor statistical testing, and neglect of differences in the route of nutrition), there is a need for randomized controlled prospective trials with exact definitions of the disease.