Minimally invasive esophagectomy is rapidly emerging as a suitable surgical alternative to the open technique. This retrospective comparative study aimed to compare two minimally invasive techniques for esophagectomy: transhiatal laparoscopy with intrathoracic or cervical anastomosis (group A) and right thoracoscopy in prone position followed by laparoscopy and left cervicotomy (group B) performed by the same surgeon (G.B.C.). The operative time, perioperative blood loss, intensive care and total hospital stays, peri- and postoperative morbidity, in-hospital mortality, number of lymph nodes dissected, and survival were the outcome measures. Between April 1999 and August 2005, 24 patients (group A) and 15 patients (group B) underwent minimally invasive esophagectomy for cancer in the authors’ department. Their median age was 61 years in group A and 61 years in group B. Preoperatively, the endoscopic location of the tumor was in the upper third in 2 cases (1 vs 1), the middle third in 11 cases (7 vs 4), and the lower third in 26 cases (16 vs 10). Two patients in each group received neoadjuvant chemo- and radiotherapy. One patient (group A) and two patients (group B) received only neoadjuvant chemotherapy, and three patients (group A) received only neoadjuvant radiotherapy. The median operative time was 300 min (range, 240–420 min) in group A and 377 min (range, 240–540 min) in group B (nonsignificant difference [NS]). The median perioperative bleeding was 325 ml (range, 100–800 ml) in group A and 700 ml (range, 100–2,400 ml) in group B (NS). The perioperative complications included one splenectomy in each group and one conversion to thoracotomy in group B. The postoperative medical complications totaled three in group A and six in group B. The postoperative surgical complications included one hemoperitoneum, one pneumothorax, five anastomotic leaks, and two recurrent laryngeal nerve paralyses in group A and two tracheal necroses, four anastomotic leaks, one colic fistula, and three recurrent laryngeal nerve paralyses in group B. The median intensive care unit (ICU) stay was 5 days (range, 2–70 days) for group A and 5 days (range, 1–180 days) for group B (NS). The median hospital stay was 12 days (range, 7–98 days) for group A and 14 days (range, 7–480 days) for group B (p = 0.05). The early mortality rate was 0%. All the specimens were free of disease. The median number of mediastinal/periesophageal lymph nodes was 3 (range, 1–10) for group A and 4 (range, 2–13) for group B (NS), and the median number of celiac/perigastric lymph nodes was 11 (range, 2–31) for group A and 10 (range, 3–22) for group B (NS). After a median follow-up period of 42.4 months (range, 2–84 months) for group A and 19.1 months (range, 1.5–34 months) for group B, 12 patients in group A died after a median period of 22 months (range, 2–55 months), and 7 patients in group B died after a median time of 15 months (range, 1.5–23 months). This retrospective comparative study showed that minimally invasive esophagectomy performed by thoracoscopy in the prone position is comparable with laparoscopic transhiatal esophagectomy in terms of the significant postoperative and survival outcomes.