We evaluated the effect of two different therapies for 23 patients with inoperable esophageal cancer, of whom 13 patients underwent placement of a metallic stent and 10 received bypass surgery. We newly defined the improved level of food intake and the effective rate based on the grade of the patient's status in order to evaluate objectively. Although the effect of stent placement was significantly better than that of bypass surgery, there were no significant differences in the improved level of food intake and the effective rate between the two groups. Four out of 10 patients in the bypass group had suture insufficiency, resulting in no oral intake and a fatal outcome. In addition, the ratio of survival period of ability to ingest orally and in-hospital mortality was significantly favorable in the stent group compared with the bypass group. In conclusion, stent placement should be the first choice for inoperable malignant stenosis or esophageal fistula.