Oesophageal cancer is one of the most fatal cancer in human. In spite of high incidence in the north of Iran and poor prognosis, there is no information regarding prognostic factors in this area. This study was conducted to determine prognostic factors of the survival of patients with oesophageal cancer under radiotherapy. We conducted a descriptive-analytical study using historical cohort that has been performed on the bases of information presented in medical record of patients with oesophageal cancer referred to the Shahid Rajaii radiotherapy center in Babolsar from 1992 to 1996. Out of 523 patients, we followed 230 patients in nonrandomized convenience sampling method based on the availability of the address for corresponding. The data was analysed using survival analysis by nonparametric method of Kaplan Meier and Cox regression model and the risk ratio (RR) of prognostic factors was determined. The results showed that the survival rates in patients with oesophageal cancer were 0.42 in 1 year, 0.21 in 2 years, 0.11 in 3 years and 0.08 in 5 years after diagnosis. The patients with the age of 50-64 ears had poorer survival rate than ones under 50 years of age and the risk ratio in this age group has been significantly increased (RR=1.73, P=0.03); the risk ratio for patients above 65 years in comparison with patients under 50 years was 1.88 (P=0.01). The females had better survival than males significantly (RR=0.71, P=0.02). For each 100 reds dose of radiotherapy, the risk ratio was decreased by 1% (RR= 0.99, P=0.05); for each session of radiotherapy, the risk ratio was decreased by 4% (RR=0.96, P=0.0001); for each square centimeter surface area under radiotherapy, the risk ratio significantly increased (RR=1.002, P=0.04). We did not observed a significant difference on survival based on the type of histology, anatomical location, and the type of treatment (P>0.05). Although the survival of patients with oeophageal carcinoma has been significantly related to sex, age, radiation dose, frequency of treatment and surface size of radiotherapy, the prognosis is extremelypoor. To compare various combined modality of treatment of oesophageal cancer, a more randomized clinical trials with enough sample size is suggested.