This paper describes the methodology of measuring the chest wall thickness using the voxel image of the Lawrence Livermore National Lab (LLNL) torso phantom. The LLNL phantom is used as a standard to calibrate a lung counter consisting of a 2 × 2 array of germanium detectors. In general, an average thickness estimated from four counting positions is used as the chest wall thickness for a given overlay plate. For a given overlay, the outer chest surface differs from that of inner one, and the chest wall thickness varies from one position to other. The LLNL phantom with chest plate and C4 overlay plate installed was scanned with a CT (computed tomography) scanner. The image data, collected in DICOM (Digital Imaging and Communication) format, were converted to the MCNP input file by using the Scan2Mcnp program. The MCNP file was visualized and analyzed with the Moritz visual editor. An analytic expression was formulated and solved to calculate the chest wall thickness by using the point detector responses (F 5 tally of MCNP). To map the chest thickness, the entire chest wall was meshed into virtual grids of 1 cm width. A source and detector pair was moved along the inner and outer surface of the chest wall from right to left at different heights from neck to abdomen. For each height (zk), (xi, yj) coordinates for the detector source pair were calculated from the visual editor and were scaled on-screen. For each (xi, yj, zk) position, a mesh thickness was measured from on-screen measurement and by solving the detector responses. The chest wall thicknesses at different positions on the outer surface of the chest were compared and verified using two methods.