We compared intraindividually the specificity of indirect pancreatic function tests before and after total (n = 4; Roux-en-Y) or subtotal (n = 6; Billroth II) gastrectomy. Before gastrectomy only 1 patient showed a falsely pathological result with the pancreolauryl test (90% specificity), while the results of all the other tests were correctly normal (100% specificity using the usual cutoff limits). After gastrectomy the respective specificities were as follows: pancreolauryl test 10%, bentiromide test 70%, fecal chymotrypsin 70%, and plasma amino acid consumption test 100%. There was no obvious difference in the reduction of specificity between subtotal and total gastrectomy. The respective preoperative to postoperative changes in the median test data were as follows: plasma amino acid consumption test +21%, bentiromide test -12%, fecal chymotrypsin -51%, and pancreolauryl test -53%. It is concluded that after gastrectomy only the plasma amino acid consumption test is unaffected by postoperative anatomic alterations.