SUMMARY Serum gastrin has been measured in 30 patients following selective gastric vagotomy. Basal serum gastrin was 52+5.7 pg/ml which was significantly lower than the corresponding level in 50 patients following truncal vagotomy (84+7.9 pg/ml). After a standard protein meal serum gastrin rose to 136±8.3 pg/ml at 60 minutes after the meal. The peak rise above basal levels was significantly lower than that achieved in patients who had undergone truncal vagotomy.These results complement our previous hypothesis that section of extragastric vagal fibres permits the release of additional gastrin above that expected with the diminution of acid secretion, and hence the decrease in inhibition of gastrin release from the antrum. Korman, Hansky, and Scott (1972) suggested that truncal vagotomy permits the release of gastrin from the antrum and other sites. They postulated that the mechanism for this release is a combination of diminution in gastric acid secretion and removal of an inhibitor to extragastric gastrin release by section of extragastric vagal fibres.The Roberts, Stedeford, and Lancaster, 1969) or postoperatively by the acid secretory response to insulin hypoglycaemia (Hollander, 1946). Only one patient of those tested had evidence of incomplete gastric vagotomy.Four patients with a complete gastric vagotomy were then further investigated. Each fasting patient was given a standard protein meal (Korman, Soveny, and Hansky, 1971) and peripheral venous blood collected at -30, 0, 15, 30, 45, 60, 75, 90, 105, and 120 minutes after protein. The protocol was identical to that reported for truncal vagotomy (Korman et al, 1972). The serum gastrin responses could thus be compared. Serum gastrin was estimated by radioimmunoassay (Hansky and Cain, 1969;Hansky, Soveny, and Korman, 1971b