Background: Intrathoracic anastomosis after oesophagectomy has recently been associated with reduced functional morbidity compared to a cervical anastomosis. Methods: From January 2011 until August 2012, all operable patients were scheduled to undergo minimally invasive oesophagectomy (MIE) with intrathoracic anastomosis. Patient characteristics, complications, morbidity and mortality were prospectively registered and analysed. Results: Forty-five patients underwent MIE with intrathoracic stapled end-to-side anastomosis. Major changes in operative technique were made 2 times due to non-satisfactory results, dividing the patients into 3 groups. One patient in group 1 died. The anastomotic leakage rate decreased from 44% in group 1 to 0% in groups 2 and 3 (p = 0.007). The pulmonary complication rate decreased from 67% in group 1 to 44% in group 2 (not significant, NS) and 22% in group 3 (p = 0.04). The median hospital stay decreased from 17 days in group 1 to 14 days in group 2 (NS) and 8 days in group 3 (p < 0.001). There were no stenoses, no dilatations and no patients with recurrent laryngeal nerve palsy. Conclusions: The introduction of the intrathoracic anastomosis led to favourable functional results but was initially associated with considerable morbidity. Results improved after changing operative techniques, but the learning curve may also be responsible.