Objectives-Concomitant radiochemotherapy improves survival from inoperable esophageal cancer compared to radiotherapy alone. Several regimens of radiotherapy (standard or concentrated split course radiotherapy) are used, however the optimum protocol remains to be determined The aim of this study was to analyze the efficacy and tolerance of concentrated concomitant split course radiochemotherapy. prognostic factors as well as those positively influencing complete response were also studied. Methods - This multicentric phase II trial looked at patients with histologically proven, inoperable, squamous cell esophageal carcinoma without metastases or invasion of the tracheobronchial mucosa. Treatment included 3 cycles of chemotherapy by 5-FU continuous infusion (800 mg/m(2).d D1-D5, D22-D26, D43-D47) cisplatin (70 mg/m(2) D2, D23, D44) and radiotherapy 15 Gy/5d (D1-D5, D22-26, D43-D47). Efficacy was analyzed by endoscopy, biopsy and computerized axial tomography during the 12(th) week of treatment. Results - The trial included 122 patients from 21 centers (110 M and 12 F, mean age 63.1 +/- 8.6 years, range 40 - 78). In accordance with the TNM-UICC classification (1978), 8 patients were classified stage I (T1 N0) 13 stage II (T2 N0), 100 stage III (T3 and/or N1) and stage was unknown in 1 patient. Median follow-vp was 63 months. Treatment was complete in half of the patients. 5 premature deaths (4.1%) were recorded over the treatment period, of which was directly linked to the toxicity of the the treatment. 16% of patients showed at least one severe side-effect. 117 patients received all 3 cycles of the treatment 88 of them without delay, and all were evaluated. 58 (47.5% of the patients included) showed a complete response with a negative biopsy, 36 (29.5%) showed a partial response, 13 (10.7%) were stable and 10 (8.2%) showed progressive disease. The median duration of complete responses wets 11.5 months. Symptomatically, dysphagia improved in 80% of the the cases, performance status in 40%, and weight gain was observed in 30% of the patients with weight loss. At evaluation, oral feeding was impossible in 4 patients only and possible in 113 patients; however, endoscopic treatment of the dysphagia remained necessary in 28 patients. Median survival in the 122 patients included was 13.0 +/- 1.6 months and survival rates were 52.9, 29.8 and 12.1% at 1, 2 and 5 years, respectively. Three pretherapeutic prognostic factors influenced survival in a multivariate analysis: initial severe dysphagia (risk of premature death increased 3-fold in the first year), circumferential extension and the differentiated nature of the tumor (risk of death doubled regardless of the time delay). Factors influencing a complete clinical response were on early tumor siege, a poorly differenciated tumor in patients older than 65, and no circumferential extension. The risk of recurrence was 54.8% at 1 year in the 58 Patients with complete remission. Complete circumferential extension and a well or moderately differentiated tumour influenced recurrence. Conclusion - This trial confirms the efficacy and good tolerance of concentrated split course radiochemotherapy in patients with inoperable esophageal cancer with a 5-year survival rate of 12%. This reinforces the need fbr a comparative trial (split course irradiation vs standard irradiation) such as the one currently being conducted in France.