It is well known that all the different procedures of vagotomy have a great variability in the clinical results and a high incidence of positivity in the Hollander's tests, this probably due to the arrangement of the vagus nerves in the low third of the esophagus and its connection with the mienteric plexus. In a group of ten dogs with a gastric fistula we obtained first a sufficient number of assays with vagal stimulation, later on vagotomy was done at the level of the esophago-gastric union (EGU), getting a decrease in the gastric secretion of hydrochloric acid of 64%, afterwards a second section of the vagus nerves was practiced, five centimeters above the EGU having an inhibition of 90% and finally 95% was attained when the vagotomy was done 15 centimeters above the EGU. It is concluded that it is needed a long vagal resection of the last ten centimeters, instead of the classic section of the nerve at the gastroesophagic union; it is the same case with a parietal cell mass vagotomy.