In 1978, postoperative radiation therapy was suggested to be an additive treatment of carcinoma of the esophagus. In 1987, we began a multicenter trial of patients with epidermoid carcinoma of the middle third or the lower third of the esophagus. Isolated surgical treatment (119 patients) was compared with surgical treatment followed by radiation therapy (102 patients). Randomization was performed according to the results of pathologic examination of all resected lymph nodes. Patients were divided into three groups: N0, no lymph node invasion; N+1, invasion of paraesophageal lymph nodes, and N+2, invasion of distal lymph nodes (recurrent, tracheal and celiac). All patients included in this trial had a curative esophageal resection. In the group receiving radiation therapy (102 patients), a total dose of 45 to 55 Grays was delivered in the mediastinal, right and left supraclavicular areas and also to the celiac area when celiac lymph node invasion was present. Two hundred and twenty-one patients were randomized and observed from three to nine years; life table method (log rank method) was used to compare survival and local recurrence. In all patients studied, the survival rate was significantly higher (p less than 0.01) in patients without lymph node invasion (group N0) than in patients with lymph node invasion (groups N+1 and N+2). Postoperative radiation therapy did not improve survival in any group. This lack of improvement in survival was present regardless of lymph node status. However, there were significantly fewer recurrences in patients receiving radiation therapy as compared with those not receiving radiation therapy.