The vagal influence on gastroduodenal content of gastrin was studied in duodenal ulcer patients. Endoscopic biopsies were assayed for total concentrations of gastrin, and fractionated for measurements of gastrin components before and after total vagotomy. Antral concentration was 17-0 +/- 2-0 mug gastrin/g mucosa (mean +/- s.e.m.) in 52 unoperated patients compared with 25-2 +/- 3-2 mug in 32 vagotomized patients. In 14 patients studied before and 3 months after vagotomy antral content was almost doubled (10-9 +/- 2-3 and 20-4 +/- 2-9 mug respectively). In the duodenal bulb the concentration was 2-4 +/- 0.3 mug in 37 unoperated patients and 2-2 +/- 0.4 mug in 19 vagotomized patients. In 6 patients in whom measurements were made throughout the duodenum, gastrin concentrations were slightly but significantly lower 3 months after vagotomy. Fractionations of pooled homogenates on Sephadex G-50 showed that gastrin component III (gastrin-17) made up 95 per cent of the antral gastrins before and after operation. In the duodenum component III constituted more than half of the gastrins preoperatively, but only one-third postoperatively. The total amount of gastroduodenal gastrin was considerably increased by vagotomy, and it is well established that in serum the gastrin concentration is also higher postoperatively. The most likely implication of these findings is that in man the vagus, directly or indirectly, suppresses gastrin production.