AIMS: Etiological investigations proposed for patients with acute pancreatitis have been evolving considerably these past few years, significantly limiting the number of cases labeled idiopathic. The aim of this study was to determine the incidence of non alcoholic non biliary pancreatitis and identify causes, comparing severity by etiology. PATIENT AND METHODS: This retrospective analysis included 108 patients managed from October 1996 to April 2005. Standar-dized extensive etiological investigations were performed. The following criteria of severity were recorded: peak CRP value, Ranson score, Balthazar score, duration of hospital stay and pseudocyst occurrence. RESULTS: The cause of acute pancreatitis was alcohol (N=45), gallstones (N=50), obstruction (N=10), unknown (N=10), drugs (N=9), auto-immunity (N=4), infections (N=3), post-operative (N=2), post-ERCP (N=2), trauma (N=1), hypertriglyceridemia (N=1), genetic (N=1). The main criteria of severity were significantly different between non alcoholic non biliary pancreatitis and the other causes (CRP>120 mg/L, Ranson score>3 and Balthazar score > or =D) while other criteria (pseudocyst occurrence and duration of hospitalisation) were similar. Mean peak CRP was 79.5 mg/L for the overall population and varied significantly by etiology: peak CRP for drug-induced acute pancreatitis (4.6 mg/L) was significantly lower than for the other causes (P<10(-6)). CONCLUSION: This study shows that non alcoholic non biliary causes account for one third of the cases of acute pancreatitis, usually with a mild to moderate presentation. As the mean peak CRP value is significantly lower in drug-induced acute pancreatitis, careful search for an adverse drug reaction is appropriate in patients with acute pancreatitis of unknown cause and a low peak CRP level.