Fifty-five patients presenting with acute abdominal symptoms and found to be hyperamylasaemic underwent early biliary tract investigation, giving 31 patients in whom the presence of gallstones was suspected. In accordance with the protocol of a randomized controlled trial of early elective biliary tract surgery for patients suspected of having acute gallstone pancreatitis, 19 of these patients underwent laparotomy at a mean of 6.9 days after emergency admission. In this group operation showed that four patients had biliary tract stones and pancreatitis; ten patients had calculous cholecystitis (53 per cent) but no stigmata of pancreatitis; four patients had pancreatitis but no stones; one had a negative laparotomy. None of this group was found to have ampullary obstruction due to an impacted stone. Biliary tract investigations carried out during the first week following admission were unhelpful or misleading in 14 out of the whole group of 55 patients, and in all of those patients (11 per cent) who died or required surgical intervention during the same hospital admission. There appears to be a pathological heterogeneity among patients diagnosed as 'gallstone pancreatitis' on clinical and biochemical grounds alone.