OBJECTIVE A reduced exocrine pancreatic function supports the diagnosis of chronic pancreatitis (CP) in symptomatic patients. A sensitive test for a reduced exocrine function is decisive, especially when morphological changes are missing. The aim of this study was to compare the indirect faecal elastase 1 (FE-1) test with the direct Lundh test in patients with and without definite diagnostic imaging findings. METHODS Eighty-nine patients with clinical signs suggesting CP or having an established diagnosis of CP had a Lundh test and an estimation of FE-1 performed. All patients underwent abdominal ultrasonography and/or computed tomography. RESULTS A significant correlation (r = 0.70, p < 0.02) was found between FE-1 and meal-stimulated intraduodenal lipase. Using the Lundh test as reference, the predictive values of a positive and negative FE-1 test were for all patients investigated 81 and 73%, respectively. Patients with equivocal imaging findings had lower predictive values (positive predictive value 57%; negative predictive value 71%) as compared with patients with moderate or marked imaging findings (positive predictive value 84%; negative predictive value 78%). Fair to moderate chance-corrected agreement was found between Lundh test and FE-1 concentration. CONCLUSIONS In patient with imaging findings suggesting CP, FE-1 determination is a highly sensitive test for exocrine pancreatic function, but in patients with equivocal imaging findings, the predictive power of FE-1 limits the test to serve as a reliable diagnostic tool.