The aim of the present study was to evaluate the safety of mediastinal microdialysis and its effi cacy regarding the early diagnosis of anastomotic leakage after gastroesophageal resection for esophageal cancer. Eight consecutive patients were included; one patient was excluded for reasons of catheter malfunction. The tip of the mediastinal microdialysis catheter was placed close to the anastomosis and held by a 4-0 absorbable suture. A subcutaneous microdialysis catheter placed in the pectoral region served as a reference. Samples collected every 4 h in the fi rst 8 postoperative days were analyzed for lactate, glucose, pyruvate, and glycerol and the lactate/pyruvate ratio (L/P ratio) was calculated. There were no procedure-related complications. Six patients had an uncomplicated postoperative course. In one patient, the L/P ratio was within normal range during the fi rst 20 h postoperatively, but a steady and signifi cant increase in L/P ratio then occurred, reaching a maximum of 105 after 124 h. The patient developed leakage symptoms on day 3, but endoscopy was unable to demonstrate any leakage. On the sixth postoperative day, blue dye administered through the nasogastric tube was recovered in the pleural drain. The leakage was treated with a covered selfexpanding metal stent; an immediate and signifi cant drop in the L/P ratio occurred. Mediastinal microdialysis seems to be a safe and promising tool in the early diagnosis of anastomotic leakage in patients undergoing gastroesophageal resection for cancer.