Esophageal cancer is still one of the most widespread diseases, and surgery for esophageal carcinoma is very stressful for patients. Even though lymph node metastasis occurs more frequently in cases of early esophageal cancer than it does in cases of gastric cancer, surgeons prefer to avoid lymph node dissection if possible, thereby subjecting patients to less invasion. Thus, the aim of the present study was to examine the possibility of predicting lymph node metastasis on the basis of tumor location, quantification theory II analysis of tumor expression of genetic markers in primary esophageal cancer. Surgical specimens from 63 patients of esophageal cancer with submucosal invasion were examined for the relationship between tumor location and lymph node metastasis. In 19 of these 63 patients, p53, p21(Waf1), and proliferating cell nuclear antigen (PCNA) were examined immunohistologically, and to quantify the risk of lymph node metastasis, computer analysis was performed on the basis of quantification theory II, in which pathological lymph node metastasis (pN) was the objective variable and "high" or "low" expression of each of the three markers was the predictive variable. Tumors located in the lower third of the esophagus had no lymph node metastasis to the upper mediastinal region, and were thus indicated for trans-hiatal esophagectomy. A coefficient greater than 0.91 predicted node negative disease accurately without false-negative results; false-positive results were obtained for 54.5% of patients with a coefficient less than 0.064. Thus, we found that quantification theory II may be useful when considering indications for surgery without lymph node dissection in some cases of T1 esophageal carcinoma.