The bentiromide test for exocrine pancreatic function was carried out in normal volunteers, patients with cystic fibrosis (CF) without clinical evidence of pancreatic dysfunction, and CF patients with clinically significant exocrine pancreatic insufficiency. The test was performed with and without the concomitant administration of a Lundh test meal. p-Aminobenzoic acid was given on a separate occasion to eliminate false positives due to factors unrelated to pancreatic disease. Correct classification of 25 CF patients with pancreatic insufficiency and 9 CF patients without clinical pancreatic dysfunction was possible by interpreting the results of the above three tests. Isoamylase determinations would have misclassified 20% of the CF patients with pancreatic insufficiency, but were able to detect the CF patients without clinical pancreatic dysfunction on the basis of an elevated pancreatic amylase isoenzyme. The bentiromide test results were normal in CF patients without clinical pancreatic dysfunction despite previous findings of decreased bicarbonate secretion in this group. However, the bentiromide test did appear to be useful in the evaluation of therapeutic intervention with exogenous pancreatic enzymes and other adjunctive therapy.