The reliability of the para-aminobenzoic acid (PABA) test (performed in the conventional manner, i.e. without control day) and of the pancreolauryl test was assayed in respect of the exocrine pancreatic capacity measured by using the secretin-caerulein test in 57 subjects, 22 of which were suffering from chronic pancreatitis. When 50 and 20% urinary excretion of the orally administered Bz-Ty-PABA and pancreolauryl, respectively, were chosen as the lower normal limits, the PABA test showed a specificity quite similar to that of the pancreolauryl test (97 and 95%, respectively) despite the lack of a control day test, but a lower sensitivity (39 vs. 83%). The association of both tests was not advantageous compared with the pancreolauryl test alone.