The secretin provocation test for gastrinoma is based on the premise that secretin decreases or has no effect on serum gastrin in normal and duodenal ulcer subjects while inducing a paradoxical rise in gastrinoma. We reexamined the serum gastrin response to pharmacologic amounts of secretin in normal volunteers (N = 17) and unoperated duodenal ulcer patients (N = 10). GIH secretin caused significant early gastrin rises from baseline in both groups (P less than 0.05). The gastrin response curves after secretin were not significantly different between normal and ulcer subjects. Similar gastrin rises were seen when synthetic secretin was administered to normal subjects (N = 6). In normal volunteers, intravenous bolus saline (N = 10) or amino acid (L-cysteine, N = 8) caused no change in serum gastrin from baseline. The gastrin response curves to GIH secretin and saline were significantly different (P less than 0.05). Our findings suggest that the gastrin rise in gastrinoma patients after secretin is an exaggeration of the normal response and not paradoxical.