The main causes of pancreatic inflammation worldwide are biliary lithiasis and alcoholism. However, 10 to 30% of patients have been considered to have "idiopathic" acute pancreatitis. Recently, some studies showed that a significant rate of the so called idiopathic pancreatitis are caused by microlithiasis and/or biliary sludge, identified by the presence of cholesterol monohidrate and/or calcium bilirubinate microcrystals in the biliary sediment. In the present study, the analysis of microcrystals from bile obtained during endoscopic retrograde cholangiopancreatography was done in patients with pancreatitis (idiopathic, biliary or alcoholic--20 in each group). Patients with idiopathic pancreatitis and microcrystals in the bile underwent cholecystectomy whenever possible. Those who refused or were inapt to surgery underwent endoscopic sphincterotomy or received continuous therapy with ursodeoxycholic acid. Patients with idiopathic pancreatitis without biliary crystals did not receive any specific treatment. The prevalence of biliary microcrystals in patients with idiopathic pancreatitis (75%) and biliary pancreatitis (90%) was significantly higher than in those with alcoholic pancreatitis (15%). In the identification of the etiology of biliary pancreatitis, the presence of microcrystals had a sensitivity of 90%, specificity of 85%, positive predictive value of 85.7%, negative predictive value of 89.4% and accuracy of 87.5%. In the patients with recurrent idiopathic pancreatitis, with biliary crystals, there was an statistically significant reduction in the number of pancreatitis episodes after specific treatment. In the follow-up of this group during 23.3 +/- 4.8 months, recurrence of pancreatitis occurred only in patients with "persistent biliary factor" (choledocholithiasis and/or persistence of cholesterol monohidrate). All patients with idiopathic pancreatitis who underwent cholecystectomy had chronic cholecystitis. Moreover, cholelithiasis was present in one case. In the ultrassonographic follow-up of the patients with idiopathic acute pancreatitis with microcrystals in the bile, cholelithiasis was detected in one case. In the subgroup of five patients with idiopathic pancreatitis without biliary microcrystals recurrence occurred in one case. Ultrassonographic study during follow-up did not reveal biliary stones in any of these patients. We concluded that the detection of biliary microcrystals in "idiopathic" pancreatitis suggested an underlying biliary etiology, even if occult. What's more, early specific therapeutic procedure (cholecystectomy, endoscopic sphincterotomy or ursodeoxycholic acid) in patients with recurrent idiopathic pancreatitis with microcrystals in the bile reduced significantly the recurrence during the follow-up. Finally, acute pancreatitis (specially recurrent) should not be called idiopathic before the microscopic analysis of the bile, aiming to detect or exclude the presence of microcrystals.