Proton radiography is an imaging technique with potential application in proton radiation therapy. The ability of a proton radiograph to differentiate anatomical features in the thoracic region, such as heart, lung, rib cage, shoulder, etc., was qualitatively investigated using Monte Carlo simulations. A patient with a stage IIIA non-small cell lung cancer tumor located in the right upper lobe and mediastinum was considered for this study. The GEANT4 Monte Carlo toolkit was used to simulate proton transport through a proton nozzle and through the lung area of the patient, registering in a phase-space file the entry and exit energy, position, and motion direction of each proton. The Monte Carlo simulation ran a total of 10 million histories with the highest deliverable energy of 235 MeV at the Francis H. Burr Proton Therapy Center. The proton radiograph was then generated independently of the Monte Carlo simulation, using a numerical algorithm to input the proton position, direction of motion, and energy kept in the entry and exit phase-space files. The proton radiograph was compared to the standard portal X-ray image for tissue and tumor contrast, and for visibility relative to the background lung tissue. The preliminary results with GEANT4 showed that the proton radiography can produce images with good spatial resolution and excellent soft tissue contrast, resulting in better tumor edge localization within a soft tissue background region such as the lung.