The present work evaluates the methodology and standards of acute hemorrhagic-necrotizing pancreatitis induced by feeding a choline-deficient, ethionine-supplemented (CDE) diet to mice. The diet model appears to be a good approximation of severe necrotizing human pancreatitis. Both the gross and histological appearance of the pancreatic and peripancreatic inflammation as well as the clinical and biochemical course of diet-induced pancreatitis resemble human disease. By limiting the period of feeding the diet, one can control the mortality at any desired level between 0 and 100%. Ascites, acidosis, hypoxia and hypovolemia occur in this model as well as in human pancreatitis. The time course of the morphological and biochemical alterations have extensively been studied and are, thus, well defined in this model. Despite the differences in pathogenesis of pancreatitis induced in this model versus human disease, the experimental pancreatitis and clinical pancreatitis share several pathophysiologic features. Therefore, the model is suitable to study pathophysiologic aspects of this disease. The diet model is particularly well suitable to study the potential for new therapeutic substances. The small size of the animals used, however, is a limitation for the evaluation of surgical procedures and of new diagnostic tools. Several pitfalls and problems have to be considered in order to obtain valuable data. The amount of injury produced by the CDE diet depends critically on sex, age and weight of the mice. Special care has to be taken to guarantee that the intake of the CDE diet is identical between different experimental groups. Therefore, each set of experiments needs to include a separate control group of mice which receive the CDE diet without any other special treatment. The potential benefit of an experimental therapy can be assessed by measuring survival, various biochemical and histological features, and alterations in hematocrit, pH and blood gases.