Purpose : To analyse clinical outcome and prognostic factors according to treatment modality, this paper report our experience of retrospective study of patients with esophageal cancer Materials and Methods : One hundred and ten patients with primary esophageal cancer who were treated in Presbyterian Medical Center from May 1985 to December 1992. We analysed these patients retrospectively with median follow up time of 28 months, one hundred and four patients($95{\%}$) were followed up from 15 to 69 months. In methods, twenty-eight patients were treated with median radiation dose irradiated 54.3Gy only. Fifty-six patients were treated with combined chemoradiotherapy. Sixteen cases of these patients were treated with concurrent chemoradiation and the other patients(forty cases) were treated sequential chemoradiotherapy. In concurrent chemoradiotherapy group, patients received 5-FU continuous IV infusion for 4 days. Cisplatin IV bolus. and concurrent esophageal irradiation to 30 Gy. After that patients received 5-FU continuous IV, Cisplatin bolus injection and Mitomycin-C bolus IV, Bleomycin continuous IV, and irradiation to 20 Gy. In sequential chemoradiotherapy group, the chemotherapy consisted of 5-FU 1,000mg/$m^2$ administered as a continuous 24 hour intravenous infusion during five days and Cisplatin 80-100mg/$m^2$ bolus injected, or Bleomycin, Vinblastine, Cisplatin, Methotrexate were used of 1 or 2 cycles. After preoperative concurrentm chemoradiation twenty-six patients underwent radical esophagectomy. Results : Ninety-three patients could be examined for response assessment, By treatment modality, response rates were $85.1{\%}$ for radiation alone group and $86.3{\%}$ for combined chemoradiation group. But in operation group, after one cycle of concurrent chemoradiation treatment, response rate was $61.9{\%}$. The pathologic complete response were $15.4{\%}$ in operation group. Overall median survival was II months and actuarial 5-year survival rate was $8{\%}$. The median survival interval was 6 months for radiation alone group, 11 months for combined chemoradiation group and 19 months for operation group. And also median survival was 19 months for complete responder group that 8 months for noncomplete responder group. In univariative analysis, statistically significant prognostic factors were tumor size, clinical stage, tumor response, and operation. In multivariative analysis, significantly better survival was associated with clinical stage, tumor response, radiation dose, and operation. Conclusion : Compared with radiotherapy alone, combined multimodality may improve the median survival in patients with localized carcinoma of the esophagus and toxicity is acceptable.