To determine if heterogeneity correction significantly affects commonly measured dosimetric parameters used to predict for pulmonary toxicity in patients who receive radiation for lung cancer. Lung cancer patients treated with 3-D treatment planning between January 1, 2000 and September 2003 were identified. Only patients receiving a minimum of 40 Gy with a minimum follow-up of 3 months were included. Patients with prior history of radiotherapy were excluded from this study. For each patient, the parameters V20, V30, V40, V50, mean lung dose (MLD), and maximum lung dose (MaxD) were measured from the heterogeneity-uncorrected plan and compared with their corresponding values on the heterogeneity-corrected plan. Those patients developing radiation pneumonitis were identified, and parameters were correlated with the risk of radiation-induced pulmonary toxicity. Sixty-eight patients met our inclusion criteria. A majority of patients had tumor involving the upper lobes. Most patients were treated with initial anteroposterior and posteroanterior (AP/PA) fields followed by off-cord obliques. On univariate analysis, all parameters correlated with radiation pneumonitis (p-values <0.05). On multivariate analysis, V20 was found to be nearly significant (p = 0.057) for predicting radiation pneumonitis. Plotting each parameter from the homogeneous plan against the corresponding heterogeneous parameter, a highly significant correlation was seen for V20, V30, V40, V50, MLD, and MaxD with all p-values <0.0001. For V20, the slope of the best-fit line is 1.00 indicating essentially no difference when corrected for heterogeneity shown in Fig. 1. Plotting similarly for MLD, the slope is 1.04. Therefore, a 4% difference is seen when correcting for heterogeneity. In addition, plotting MLD against V20 shows a highly significant correlation. The implication is that either parameter may be used in predicting for risk of pneumonitis. When analyzing dosimetry of the normal lung and the risk of developing pneumonitis, parameters measured from heterogeneity-corrected plans highly correlate with those from uncorrected plans. Therefore, normal lung DVH’s are not significantly affected by heterogeneity correction. Either V20 or MLD may be used to assess pneumonitis risk.