BACKGROUND The Lundh test is a usual means of estimating the enzyme secretory capacity of the gland. During this procedure, however, a major proportion of the test meal is removed from the duodenum together with the gastric, duodenal, and pancreatic secretions and the bile. This study was undertaken to compare the pancreatic enzyme secretion induced by the Lundh procedure with that resulting from stimulation of the normal digestive process, by reinfusion of the aspirated duodenal juice. METHODS Nine men (mean age: 46.7, range 42-55 yr) free from pancreatic disease were studied. Pancreatic secretion was measured via a multiple lumen tube by aspiration of the duodenal juice. After a basal period the Lundh test meal was placed in the stomach and the duodenal juice was completely aspirated. On a separate day, the procedure was repeated, but the aspirated duodenal juice was reinfused into the upper jejunum. RESULTS In the first 30 min of the test period, the enzyme outputs were the same on both test days. In the 30-60-min period, the lipase output, and in the 75-90-min period, the amylase output was significantly lower during the Lundh test compared with the jejunal reinfusion test. The CCK levels were significantly above the basal level at 20 and 40 min, but the increase was significantly lower during the traditional Lundh test. No significant difference in gastrin release was observed during either the Lundh or the reinfusion test. CONCLUSIONS In the traditional Lundh test, the trypsin secretory capacities of the gland are measured appropriately, but the lipase and amylase secretory capacity and the CCK release are not fully represented compared with the reinfusion test. An association between the lower CCK release and lipase amylase secretion is suggested.