The authors report 4 cases of esophageal fistula following a total gastrectomy. One patient died after a reoperation for a subphrenic abscess, another just before a coloplasty 3 months after an esophageal exclusion. Two patients have been successfully treated by an operation which comprised the removal of the fistula and an intra-thoracic esophago-jejunal anastomosis. This procedure, albeit risked, is probably a better option than the esophageal exclusion usually recommended, particularly in the patients with a malignant disease who have a short life expectancy.