Acute pancreatitis is a severe disease with high mortality. Clinical studies can bring some data about etiology, pathogenesis and the course of acute pancreatitis. However, studies concerning early events of this disease and the new concepts of treatment cannot be performed on humans, due to ethical reasons. Animal models of acute pancreatitis have been developed to solve this problem. This review presents currently used experimental models of acute pancreatitis, their properties and clinical relevance. Experimental models of acute pancreatitis can be divided into in vivo (non-invasive and invasive) and ex vivo models. The onset, development, severity and extent of acute pancreatitis, as well as the mortality, vary considerably between these different models. Animal models reproducibly produce mild, moderate or severe acute pancreatitis. One of the most commonly used models of acute pancreatitis is created by administration of supramaximal doses of cerulein, an analog of cholecystokinin. This model produces acute mild edematous pancreatitis in rats, whereas administration of cerulein in mice leads to the development of acute necrotizing pancreatitis. Acute pancreatitis evoked by retrograde administration of sodium taurocholate into the pancreatic duct is the most often used model of acute severe necrotizing pancreatitis in rats. Ex vivo models allow to eliminate the influence of hormonal and nervous factors on the development of acute pancreatitis.