PURPOSE: To analyze the association of patient- and treatment-related factors with the onset of radiation pneumonitis in a homogeneously treated cohort of patients suffering from small cell lung cancer (SCLC). PATIENTS AND METHODS: 242 patients with SCLC staged as limited disease, who had been treated with chemotherapy and three-dimensional conformal radiotherapy, were retrospectively analyzed. Pneumonitis was defined by typical symptoms and radiographic findings and judged clinically relevant, if drug administration and hospitalization were necessary. Patient- (age, gender, smoking history, performance status, tumor localization, benign lung disease) and treatment-related parameters (V(10)-V(40), mean lung dose [MLD]) were analyzed using chi(2)-tests for categorical parameters and logistic regression for continuous variables. RESULTS: 33 patients (13.6%) developed a clinically relevant pneumonitis, of whom three patients died. All cases of pneumonitis developed within 120 days. None of the patient-related parameters correlated significantly with the onset of pneumonitis. Considering treatment-related parameters, a significant correlation of V(30) in regard to total lung and V(40) in regard to ipsilateral, contralateral and total lung to the risk of pneumonitis was found. So, the estimated risk of a clinically relevant pneumonitis increased from 10% given a V(30) of 13% to 30% given a V(30) of 35%. In contrast, no significant correlation was found for V(10) and V(20) and only a trend for MLD. CONCLUSION: In this series, high-dose radiation volume parameters, i.e., V(30) and especially V(40), were identified as the most important factors for the development of radiation pneumonitis. Low-dose radiation volume parameters and clinical parameters played an inferior role in predicting the pneumonitis risk.