Purpose: To assess the efficacy of intensity modulated radiotherapy (IMRT) optimization for lung cancer based on dose‐to‐structure mass objectives. Methods: Three patient cases are planned for two radiotherapy deliveries; each optimized using either dose‐volume histogram (DVH) constraints or dose‐mass histogram (DMH) constraints. An inhale‐phase optimization simulates breath hold (BH) treatment; an average CT (aCT) optimization simulates treatment under free breathing. DVH plan objectives include 70 Gy to 98% of the PTV while minimizing lung V20, esophagus V25, heart V30. DMH optimization utilizes the same beam angles with corresponding objectives; 70 Gy to 98% of the PTV mass and minimum dose to relative mass of risk structures. Relative mass at objective dose levels are compared for DVH and DMH optimization. Results: For BH plans, DMH optimization maintains or improves target‐mass coverage (up to 2% increase in mass at 70 Gy) and decreases lung mass by up to 2%, heart mass by up to 4%, and esophagus mass by up to 2% at the objective dose levels. The DMH plans optimized on aCT increase target mass coverage by <:1% in all cases and do not improve risk structure mass sparing compared to DVH based plans. The DVH to DMH optimization differences are patient and radiation path dependent. For a complete set of voxels composing a given volume, irradiated voxels are a subset which can have a different mean density than the entire set. This leads to variations in DVH and DMH solutions for the PTV, lungs, esophagus, and heart. Preferred (reduced) radiation path‐lengths through regions of lower density lung are observed in BH‐DMH optimization. Conclusions: DMH optimization has the potential to improve the therapeutic ratio on well‐defined geometry (for example during BH‐RT). This may have advantages in the heart, however, functional lung sub‐units must be further investigated. Support: P01CA11602 and Philips Medical Systems.