We assessed the clinical usefulness of the intraductal secretin test in order to ascertain whether it can substitute for the conventional duodenal secretin test. Duodenal juice was obtained with a triple-lumen tube and pure pancreatic juice was obtained by retrograde cannulation of the main pancreatic duct using a duodenofiberscope. Pancreatic secretion was stimulated by a bolus intravenous injection of secretin (100 units). The two tests showed comparable interindividual coefficients of variation, significantly good correlations, and comparable diagnostic efficiencies. The intraductal secretin test showed no less reproducibility than that of the duodenal secretin test as reported in the literature. In the intraductal secretin test, secretory volume, peak flow rate, bicarbonate output, and lipase output yielded the best diagnostic efficiency, followed by amylase output and maximal bicarbonate concentration. In the intraductal secretin test, a 10-min collection provided as much information as a 20-min collection. We conclude, therefore, that the 10-min intraductal secretin test is as useful as the conventional duodenal secretin test in assessing exocrine pancreatic function and that the most discriminatory parameters are secretory volume, bicarbonate output, and amylase (or lipase) output.