Thirty-three patients with chronic pancreatitis were studied in an effort to correlate release of gastrointestinal hormones (GIH) with the degree of pancreatic insufficiency. A prospective examination was conducted of fat-stimulated release of pancreatic polypeptide (PP), cholecystokinin (CCK), and neurotensin. Seventy-two-hour fecal fat determination, endoscopic retrograde pancreatography (ERP), and the bentiromide-PABA test were used to correlate the clinical stage of disease. The ERP was classified as positive only if the changes were advanced (or "marked") according to the Cambridge Classification. Five patients were defined to have mild disease, 13 moderate, and 15 severe. Any patient with clinical evidence of chronic pancreatitis and ERP changes that were less than advanced and had normal fecal fat and bentiromide tests received a grade of mild. Patients with one abnormal test were graded moderate, and those with two or three abnormal results were graded severe. In the 33 patients, the integrated 60-minute release of pancreatic polypeptide (PP) was 37.4 +/- 6.1 ng-60 min/ml in those five patients with mild disease, 102.3 +/- 10.3 ng-60 min/ml in the 13 patients with moderate disease, and 7.6 +/- 2.2 ng-60 min/ml in the 15 patients with severe disease. The integrated 60-minute release of neurotensin was 3.8 +/- 0.4 ng-60 min/ml in mild disease, 2.0 +/- 0.3 ng-60 min/ml in moderate disease, and 0.2 +/- 0.1 ng-60 min/ml in severe disease. CCK release did not correlate with the severity of disease. Enhanced release of PP appeared to correlate well with moderate stage of chronic pancreatitis, and depressed PP release with severe disease. Stimulated levels of PP and neurotensin appear to be useful in the diagnosis and staging of chronic pancreatitis. It is concluded that measurement of fat-stimulated release of PP and neurotensin may be useful to assess severity of disease in patients with chronic pancreatitis.