Residency training programs commonly emphasize a single technique of esophagectomy, as the safety and the efficacy of teaching or performing more than one type of esophagectomy are unclear. Between 1986 and 1992, 248 patients were explored for possible esophageal resection. Thoracic surgical residents or fellows performed major components of all resections. Two hundred twenty-one patients (adenocarcinoma, 146; squamous cell carcinoma, 72; and other, 3) underwent transthoracic esophagectomy (n = 134), transhiatal esophagectomy (n = 42), or total thoracic esophagectomy (n = 45), a resectability rate of 89.1% (221/248). Complications occurred in 75% of patients with transthoracic esophagectomy, in 69% with transhiatal esophagectomy, and in 80% with total thoracic esophagectomy. The overall operative mortality rate was 6.8% (15/221). Patients with a cervical anastomosis had a higher leak rate (13%) than those with an intrathoracic anastomosis (6%). Median survival was 22 months (19% 5-year survival) and did not differ by operation type or stage. No patient with unresectable disease (n = 27) survived longer than 10 months. Survival for patients with adenocarcinoma stages 3 and 2a suggested a trend toward improved survival after transthoracic esophagectomy despite similar rates of local and distant recurrence. Transthoracic esophagectomy, transhiatal esophagectomy, and total thoracic esophagectomy performed within a residency training program have similar morbidity, mortality, and recurrence rates as those in other modern series. A specific technique of esophagectomy can be selected for individual patients. Survival and sites of recurrence primarily reflect disease stage, not the technique of esophagectomy used.