BACKGROUND A simple, rapid test is specific and sensitive enough to distinguish, in patients with clinically suspected acute pancreatitis, those whose abdominal pain is indeed of pancreatic origin has proved elusive. METHODS In two consecutive series of surgical patients in a teaching hospital, whose acute abdominal pain turned out to be due to acute pancreatitis (n-57) or extrapancreatic in origin (n=40), we studied urinary trypsinogen-2 in two ways. A test strip, incorporating monoclonal antibodies to two epitopes on trypsinogen-2, recorded a blue line when concentrations exceeded 50 microgram/L; we also measured trypsinogen-2 concentrations in the laboratory. RESULTS In the patients with acute pancreatitis the test strip was positive in 52 and negative in five, whereas in the 40 extrapancreatic controls there were four false positives. In a further set of 57 orthopaedic controls, one urine was strip-test positive. Concentrations of urinary trypsinogen-2 and the test-strip results were in good agreement and in only three of the 154 patients were the two approaches discrepant, at the 50 microgram/L cut-off. CONCLUSIONS These findings, in patients whose acute abdominal pain was known to be pancreatic in origin or not, are encouraging but need to be confirmed in a consecutive series of patients in whom the diagnosis of pancreatitis is in doubt.