OBJECTIVE: To explore the clinical and physical factors that might give rise to radiation-induced esophagitis in three-dimensional conformal radiotherapy for non-small cell lung cancer. METHODS: To collect the clinical and physical records and follow-up information of 106 NSCLC patients without undergoing surgery in our hospital. χ(2) test, linear tendency test and analysis of variance were employed to analyze the relationship between occurrence of radiation-induced esophagitis and clinical and physical treatment. Logistic analysis was also used for multivariate analysis. RESULTS: Among the 47 cases of radiation-induced esophagitis, 31 cases were of grade I, 11 of grade II, 5 of grade III, and with a total occurrence rate of 44.3% (47/106). Radiation-induced esophagitis was correlated with Karnofsky scores, radiation sensitization and tumor location (χ(2) = 11.30, 8.45, 7.67, P < 0.05). Radiation-induced esophagitis was correlated with the length of irradiated esophagus and average dose of irradiated esophagus (F = 20.82, 83.08, P < 0.001). With the increase of the irradiated volume percentage from V20, V30, V40 up to V50, the occurrence rate of radiation-induced esophagitis was also increased, almost with a linear trend (P < 0.05). Application of all the above factors to logistic model indicated that radiation sensitization,length of irradiated esophagus, average dose and V50 were all statistically significant foactors in the occurrence of radiation-induced esophagitis (OR = 0.321, 2.850, 7.307 and 8.558, P < 0.05). CONCLUSIONS: Radiation sensitization,length of irradiated esophagus, average dose of irradiated esophagus and V50 are independent factors in the occurrence of radiation-induced esophagitis. V50 is of greater importance in the judgement of occurrence of radiation-induced esophagitis.