Abstract: The difference between two images obtained from two flash radiographic imaging experiments under the same conditions reflects the repeatability of imaging. The optical density curves with equal optical path obtained from the images are measured and linearly fitted to evaluate the imaging repeatability of an imaging system. The experiments are conducted to measure the repeatability of the radiographic system with and without back cone respectively, and the imaging repeatability of the film in the system is measured. The results show that the imaging repeatability of the system is about 5%, and the back cone shows no obvious influence on the repeatability. The repeatability error introduced by the film is about 1.6%.