Neutron radiography (NR) is a powerful non-destructive testing technique widely used, either on its own or as complementary to X-ray radiography. This technique is used for analyzing the objects in order to determine structural defects. In comparison with X-ray or g -ray radiography, X-rays are attenuated by large atomic number materials therefore they are not suitable for low atomic number materials. On the other hand, neutrons can be attenuated by low atomic number materials such as materials containing hydrogen. Three main components of neutron radiography system are source, collimator, and detector. Various effective parameters on the image quality are needed to be studied for neutron radiography. In this work, an electron accelerator-driven neutron radiography system, based on rhodotro TT200, has been simulated utilizing MCNPX code. To design an optimized photon and photoneutron target, different materials in different thicknesses and radii have been investigated. Between tungsten and tantalum, tungsten cylinder with 0.175 cm thickness and 1.2 cm radius has been chosen for the photon target. Among beryllium, deuterium oxide, tantalum and gold, a beryllium cylinder placed around photon target with 3 cm thickness and 15 cm radius has been selected as a photoneutron target due to its high neutron flux. The neutron flux of the designed photoneutron target obtained 1.8E12 n/s. The use of 12 cm Pb as a reflector around photoneutron target was investigated and the neutron flux was calculated. According to this calculation, neutron flux increased 37% in the outlet surface. In the next step, different geometries and materials were investigated in order to choose a suitable moderator. Among polyethylene, plexiglass, BeD2, and BeO, conical polyethylene with 15 cm thickness has been chosen to be placed before a collimator. This collimator is a convergent-divergent one which the converge section contains Fe and Fluental as fast neutron filters and the divergent section of this collimator contains polyethylene. Bi is also used around it as a reflector. By placing 15 cm polyethylene as moderator, Neutron radiography parameters (TNC , Thermal Neutron Content, and n/g ratio , and thermal neutron flux) were calculated. The results obtained from this work are compared with 40 MW reactor and standard ranges of neutron radiography. For instance, calculated thermal neutron flux was 6.69E4 n/cm 2 .s while the obtained one in reactor wa 8.10E4 n/cm 2 .s , and the standard range in accelerator i 10 4 -10 5 n/cm 2 .s. The obtained n/g ratio was 1.05E7 n/cm 2 .mR whereas in reactor it was 5.01E7 n/cm 2 .mR, and the standard range is more than 10 6 n/cm 2 .mR. TNC shows a better rank in this design and was obtained 74.9% while in the reactor TNC was 72%. So, the designed photoneutron source can be a suitable substitute instead of reactor source which is not economical and it shows that the results obtained using MCNP simulations are really in good agreement with the results obtained from 40 MW reactor and the standard ranges of neutron radiography.